Skip to main content

Full text of "The Safe Use Of Ultrasound In Medical Diagnosis, 3rd Edition"

See other formats


THE SAFE USE OF 
ULTRASOUND IN 
MEDICAL DIAGNOSIS 



Edited by Gail ter Haar 


3rd Edition 









The Safe Use of Ultrasound 
in Medical Diagnosis 

3rd Edition 

Edited by Gail ter Haar 



We should like to acknowledge the support of the British Medical Ultrasound Society, the 
European Federation of Societies for Ultrasound in Medicine and Biology, and the National 
Physical Laboratory (UK). Without their generosity this revision would not have been possible. 



The British Institute of Radiology 

36 Portland Place, London W1B 1 AT, UK 

www.bir.org.uk 

Published in the United Kingdom by The British Institute of Radiology 
© 1991 The British Institute of Radiology 

© 2000 The British Medical Ultrasound Society & The British Institute of Radiology 
© 2012 The Authors 

Hbir|open 

©This book is licensed under a Creative Commons Attribution-NonCommercial- 
NoDerivs 3.0 Unported License 

Some rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any 
form or by any means, electronic, mechanical or photocopying, recording, or otherwise for commercial purposes, 
or altered, transformed, or built upon, without the prior written permission of the British Institute of Radiology 

First published 1991 (978-0-905749-28-0) 
Second edition 2000 (978-0-905749-42-6) 
Third edition 2012 (978-0-905749-78-5) 

British Library Cataloguing-in Publication data 

A cataloguing in record of the publication is available from the British Library 

ISBN 978-0-905749-78-5 (print) 
ISBN 978-0-905749-79-2 (eBook) 

A print version of this book can be purchased from the BIR website 

The British Institute of Radiology has no responsibility for the persistence or accuracy of URLs for external or third- 
party internet websites referred to in this publication, and does not guarantee that any content on such websites is, 
or will remain, accurate or appropriate 

All opinions expressed in this publication are those of the respective authors and not the publishers. The publishers 
have taken the utmost care to ensure that the information and data contained in this publication are as accurate 
as possible at the time of going to press. Nevertheless the publishers cannot accept any responsibility for errors, 
omissions or misrepresentations howsoever caused. All liability for loss, disappointment or damage caused by 
reliance on the information contained in this publication or the negligence of the publishers is hereby excluded 




Contents 



Contributors 
Preface 
Chapter 1 

Chapter 2 

Chapter 3 
Chapter 4 
Chapter 5 
Chapter 6 
Chapter 7 
Chapter 8 
Chapter 9 
Chapter 10 

Chapter 11 



Glossary 
Index 



Introduction 

Gail ter Haar 

The propagation of ultrasound through tissue 

Francis A. Duck 

The acoustic output of diagnostic ultrasound scanners 

Adam Shaw and Kevin Martin 

Ultrasound-induced heating and its biological consequences 

Charles C. Church and Stanley B. Barnett 
Non-thermal effects of diagnostic ultrasound 

J. Brian Fowlkes 

Radiation force and its possible biological effects 

Hazel C. Starritt 

Bio-effects — cells and tissues 

Gail ter Haar 

The safe use of contrast-enhanced diagnostic ultrasound 

Douglas L. Miller 

Epidemiological prenatal ultrasound studies 

Kjell A. Salvesen 

Safety standards and regulations: the manufacturers' 
responsibilities 

Francis A. Duck 

Guidelines and recommendations for the safe use of diagnostic 
ultrasound: the user's responsibilities 

Gail ter Haar 



IV 

v 
1 

4 

18 
46 
69 
81 
91 
105 
125 
134 

142 



159 
163 



Hi 



Contributors 



Dr Stanley B. Barnett, MSc, PhD 

1 1/147 Darley St. West, Mona Vale, NSW 2103, Australia 
E-mail: shirlstan2004@yahoo.com 

Dr Charles C. Church, MSc, PhD 

National Center for Physical Acoustics, University of Mississippi, MS 38655, USA 
E-mail: cchurch@olemiss.edu 

Professor Francis A. Duck, PhD, DSc 

3 Evelyn Rd, Bath BA1 3QF, UK 
E-mail: f.duck@bath.ac.uk 

Professor J. Brian Fowlkes, PhD 

Department of Radiology, University of Michigan, Medical Science I, 1301 Catherine, 
Room 3226C, Ann Arbor, Ml 48109-5667, USA 

Department of Biomedical Engineering, University of Michigan, 3315 Kresge Research Building III, 
204 Zina Pitcher Place, Ann Arbor, Ml 481 09-0552, USA 
E-mail: fowlkes@umich.edu 

Dr Kevin Martin, BSc, PhD, FIPEM 

Department of Medical Physics, University Hospitals of Leicester, 
Infirmary Square, Leicester LE1 5WW, UK 
E-mail: kevin.martin42@btinternet.com 

Dr Douglas L. Miller, PhD 

Basic Radiological Sciences Division, Department of Radiology, University of Michigan SPC 5667, 
3240A Medical Science Building I, 1301 Catherine Street, Ann Arbor, Ml 48109, USA 
E-mail: douglm@umich.edu 

Dr Kjell A. Salvesen, MD, PhD 

Department of Obstetrics and Gynaecology, Clinical Sciences, Lund University, 
Box 117, SE-221 00 Lund, Sweden 
E-mail: pepe.salvesen@ntnu.nu 

Mr Adam Shaw, BA, MA (Cantab) 

Acoustics and Ionizing Radiation Division, National Physical Laboratory, 
Hampton Road, Teddington TW1 1 OLW, UK 
E-mail: adam.shaw@npl.co.uk 

Dr Hazel C. Starritt, PhD 

Medical Physics and Bioengineering, Royal United Hospital, Combe Park, Bath BA1 3NG, UK 
E-mail: hazelstarritt@nhs.net 

Dr Gail ter Haar, MA, PhD, DSc 

Institute of Cancer Research, 15 Cotswold Road, Belmont, Sutton SM2 5NG, UK 
E-mail: gail.terhaar@icr.ac.uk 



/V 



The Safe Use of Ultrasound in Medical Diagnosis 



Preface 



It is an oft observed fact that safety sessions at congresses are seldom well attended, and 
that the sneaky insertion of a lecture on a safety-related topic into a specialist session may 
be regarded by some as the opportunity for a coffee break, but the fact remains that the safe 
use of diagnostic ultrasound is the responsibility of the person conducting the scan. In order 
for appropriate judgements to be made, the practitioner must be knowledgeable about 
the hazards and risks involved in performing an ultrasound examination, and this book 
aims to provide this basic knowledge. Leading world experts in the fields of ultrasound 
physics, biology, standards and epidemiology have contributed chapters, written at a level 
that is intended to be accessible to everyone, whatever their background. Each chapter is 
extensively referenced to allow readers to delve deeper into a topic of interest if they so wish. 

Ultrasound has an unprecedented safety record, but that does not mean that we can be 
cavalier about its use. What is evident from the information presented in this book is that there 
are many gaps in our knowledge about ultrasound safety. Many of the studies on which we 
base our information and recommendations have been carried out in animal models whose 
relevance to the human is not fully understood, ultrasound exposure conditions which have 
little relevance to diagnostic ultrasound pulses, or on scanners that are no longer in common 
clinical use. While this is useful information, it must always be interpreted with care. 

It must be remembered that "absence of evidence of harm is not the same as absence of 
harm" (Salvesen et ah, 2011). It is never possible to prove a negative, all we can do is to 
use increasingly more sensitive tests and assays. It is for these reasons that professional 
societies continue to support committees whose remit is to inform and educate users about 
the safe of ultrasound, so that ultrasound imaging can continue to enjoy its reputation as 
a technique whose benefits far outweigh any potential risk. 

The publication of the third edition of this book would not have been possible without 
the generous support of the British Medical Ultrasound Society, European Federation of 
Societies for Medical Ultrasound and the National Physical Laboratories to whom I am 
extremely grateful. 

Gail ter Haar 
London, November 2012 

Reference 

Salvesen KA, Lees C, Abramowicz J, Brezinka C, ter Haar G, Marsal K. 2011. Safe use of 
Doppler ultrasound during the 11 to 13 + 6-week scan: is it possible? Ultrasound Obstet 
Gynecol, 37, 625-628. 

v 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 1 

Introduction 

Gail ter Haar 

Institute of Cancer Research, Sutton, UK 



The decision by the British Medical Ultrasound Society (BMUS), the European Federation 
of Societies for Ultrasound in Medicine & Biology (EFSUMB) and the UK National 
Physical Laboratory (NPL) to sponsor the revision of this publication on the topic of the 
safety of diagnostic ultrasound in medical practice at this time is entirely appropriate. 
In England alone, over two and a half million obstetric ultrasound scans (about four 
for every live birth) are performed every year (Department of Health, 2012). Many of 
these are carried out using the new generations of ultrasound scanners, which have the 
potential to produce significantly higher acoustic outputs than their predecessors (see 
Chapter 3). Ultrasound imaging has become more sophisticated and new techniques such 
as tissue harmonic imaging, pulse coding and contrast-enhanced imaging are becoming 
more common, bringing with them not only increased diagnostic capabilities, but also 
uncharted waters as far as safety considerations are concerned. This is not unusual; we 
have a track record of safety studies lagging behind clinical applications— there are, 
for example, no epidemiological studies concerned with the use of pulsed Doppler 
techniques. This state of affairs is not to be condoned, and there is now considerable 
effort being put into understanding the way in which an ultrasonic beam interacts with 
tissue in terms of its heating potential, and the probability of inducing mechanical effects 
such as acoustic cavitation, so that there is more chance of predicting and preventing the 
occurrence of an unwanted bio-effect. 

During the early 1990s a change was made by the Food and Drug Administration (FDA) 
in the USA that has affected all those using ultrasound for medical diagnosis. Output 
levels had been set in the 1980s simply on the basis that such conditions had been in 
use before, with no evidence of hazard. The change allowed intensities previously 
reserved only for peripheral vascular studies to be used for all studies, including first- 
trimester scanning. No epidemiological or other evidence was then, or is now, available 
to support the assertion of safety at these higher exposures. The FDA change resulted 
in the widespread availability of high specification pulsed Doppler and Doppler 
imaging modes for uses in addition to cardiovascular applications, including obstetrics. 
Recognizing the difficulty of establishing resilient safety management for this change, the 
FDA decided to pass the responsibility for safe management to the user. Manufacturers 

1 



1 Introduction 



are now able to use higher exposures than before, provided that the equipment displays 
"safety indices". These, the thermal index (TI) and the mechanical index (MI), have 
been designed to inform the user of conditions which might give rise to safety concerns 
during any scanning session. For those using ultrasound equipment, these changes in 
philosophy are of central importance to their clinical practice. The management of safety 
has become a partnership between manufacturers, whose responsibility it is to design 
and make safe equipment, and the users whose responsibility it is to understand how to 
operate the equipment safely. The primary purpose of this book is to inform users about 
the principles and evidence on which this safe practice depends. 

Two biophysical mechanisms, heating and cavitation, have become central to safety 
judgements. In order to assist those using diagnostic ultrasound equipment to make their 
own judgements on safety, the two safety indices mentioned above were introduced. The 
TI gives an approximation to the greatest temperature rise which could occur in exposed 
tissue. This tissue warming (a more realistic word to describe what may happen than 
"heating") results from the energy deposited in the tissue by ultrasound absorption. The 
highest local temperatures occur in bone in vivo, since this tissue absorbs the ultrasound 
waves most strongly. The theory for MI describes the resonant behaviour of gas bubbles 
in liquids, which could cause damage from "inertial cavitation". Gas bodies are essential 
precursors to this process and there is no experimental evidence that inertial cavitation 
occurs at diagnostic ultrasound levels in their absence. However, there are two situations 
in vivo where gas bodies may be exposed to diagnostic ultrasound. These are during the 
use of gas-bubble ultrasound contrast agents, and when ultrasound exposes tissue which 
naturally contains gas, such as the lung or intestines. These are discussed in Chapters 5 
and 8. 

When considering the safe use of ionizing radiation, the use of the ALARA (as low as 
reasonably achievable) principle is widespread and entirely appropriate. It is often 
brought up in the context of the safety of ultrasound exposures. Here it should be used 
with caution. If the assumption is correct that heating and cavitation are the two prime 
mechanisms by which hazardous bio-effects can be brought about, then, at exposure 
levels that lie below the thresholds for their occurrence (see Chapters 4 and 5) there is no 
reason for keeping exposures low, provided these thresholds are not exceeded. However, 
where exposure levels have the potential to move above the threshold then it is entirely 
appropriate to invoke the ALARA principle in an attempt to minimize potential hazard. 
At exposures below the thresholds, the risk/benefit judgement depends on uncertainties 
about the validity of these thresholds, and also about uncertainties of the existence and 
effects of other bio-effects mechanisms. 

A problem that has bedevilled the study of ultrasound bio-effects is the lack of a consistent 
method of describing "dose" . There are no separate units to describe the level of ultrasound 
exposure incident on tissue (kerma would be used to describe this aspect of an X-ray beam) 
and the ultrasound "dose" to the tissue (here units of Gray are used for X-rays). A problem 
arises in ultrasound dosimetry, with ultrasound fields being described in terms of pressure 
or intensity, neither of which give a measure of energy deposition. Either "free-field" or in 
situ values are given. In situ values have been "derated" to account for tissue attenuation 

2 



Introduction 1 



(see Chapters 2 and 3). Often, the precise nature of the parameter quoted in the published 
bio-effects literature is not given. This situation has led to problems of interpretation 
of much of the early safety literature in terms of its relevance to diagnostic ultrasound 
exposures. However, more rigour is now being applied (and, increasingly, required by 
professional journals) and we can look forward to more clinically relevant safety studies 
coming out of research laboratories. 

The intended readership of this book includes all clinical users of diagnostic ultrasound, 
including sonographers, radiologists and obstetricians, together with those using 
ultrasound in other clinical areas such as general practice, cardiology and vascular 
studies. It is also intended to provide fundamental information about ultrasound safety 
to those in clinical training. In addition, the book should be of value to clinical and 
research scientists engaged in the development of new ultrasound diagnostic methods. 
The book has been structured to aid interpretation of the "on-screen" labelling which is 
now used very widely on ultrasound scanners (see Chapters 4-6), to inform the user of 
the current status of bio-effe cts research (see Chapters 7-9); and to review the regulations 
and recommendations regarding use of diagnostic ultrasound (see Chapters 10 and 11). 

The BMUS and EFSUMB have Safety Committees. One of the functions of these Groups 
is to ensure that their members are kept informed about issues of safety. This book arose 
originally, in part, as a result of an awareness of this responsibility. This revision has 
been co-sponsored by BMUS, EFSUMB and NPL. Another effective vehicle for circulating 
and updating safety information is the internet. The websites of the BMUS and EFSUMB 
Safety Committees provide a valuable resource containing safety statements, tutorial 
articles and literature reviews. The American Institute for Ultrasound in Medicine (AIUM) 
also publishes safety related information on their Website (www.aium.org), as does 
the World Federation for Ultrasound in Medicine & Biology (WFUMB; www.wfumb.org). 

Ultrasound has an enviable record for safety. Nevertheless, modern scanners are capable 
of warming tissue in vivo, applying stress to tissue and, under some circumstances, 
damaging fragile structures adjacent to gas. It is essential that in the enthusiastic search 
for greater diagnostic efficacy the pre-eminent place gained by ultrasound as a safe 
diagnostic mode is not prejudiced. It is the responsibility of all those engaged in the 
diagnostic use of ultrasound to ensure that this is so. 

Acknowledgement 

This chapter is a revised version of Chapter 1 in the second edition. The contribution of 
Francis Duck to that chapter is acknowledged. 

Reference 

Department of Health. 2012. http://www.dh.gov.uk. 



3 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 2 

The propagation of ultrasound 
through tissue 

Francis A. Duck 

University of Bath, Bath, UK 



Summary 

• Ultrasonic waves in the frequency range 1-20 MHz are widely used for medical 
diagnostic applications. 

• Exposure is usually given in terms of peak rarefaction pressure, total acoustic power 
and acoustic intensity. 

• In situ exposure may be estimated using simple tissue models. 

• The two main bio-effects mechanisms are heating and mechanical processes. 

• The most likely tissues to experience heating are bone and adjacent soft tissues. 

• The most likely tissues to experience mechanical damage are those adjacent to gas: at 
the lung surface, in the intestine and with contrast agents. 

• Non-linear acoustic effects are particularly significant during propagation through 
fluids such as water and amniotic fluid. 



2.1 Introduction 



Ultrasound 
describes 
mechanical 
waves above 
20kHz 



Frequencies 
between 
1MHz and 
20MHz are 
used for 
diagnostic 
ultrasound 



The term ultrasound describes a mechanical wave, similar in character to audible sound, 
but at frequencies greater than 20 kHz, or 20,000 cycles per second. For medical applications 
frequencies typically above 1MHz are used. These are at least 100 times more rapid than 
the oscillations that can be detected by the ear. In this chapter, a description is given of the 
way in which waves of this frequency travel through the body, emphasizing those aspects 
that may be important when making considering judgements about the safe management 
of diagnostic uses of ultrasound. 

Particular emphasis will be given to the propagation characteristics in the frequency 
range between 1MHz and 20 MHz. At such frequencies, practical use is made of these 
waves in clinical medicine for diagnostic, therapeutic and destructive purposes, and 
therefore their propagation characteristics are of particular interest and have been most 
fully studied. From a knowledge of the wave velocities and of the degree to which tissues 

4 



The propagation of ultrasound through tissue 2 



absorb, scatter and reflect ultrasound, it is possible, in principle, to predict the manner 
by which ultrasound propagates within, and interacts with, the body. This chapter has 
two parts. In the first, a general overview is given of ultrasonic wave propagation, and 
of the properties of body tissues that affect it. In the second, this knowledge is used to 
describe what may happen to a pulse of ultrasound as it travels into tissue, so setting the 
biophysical basis for the later discussions of ultrasound safety. 



2.2 Ultrasound wave propagation 



Ultrasound is propagated in a manner identical to that of audible sound, through 
the displacements of the molecules constituting the medium in which the wave is 
travelling. It is thus a fundamentally different wave phenomenon from electromagnetic 
waves such as radio waves, infrared radiation and X-rays. The ultrasonic wave may 
propagate in the same direction as the displaced particles, in which case it is called a 
longitudinal compressional wave. Alternatively the particles may oscillate transversely, 
perpendicularly to the direction of propagation. Such a wave is termed a transverse or 
shear wave. Though shear waves can propagate in solids, and may therefore travel in 
calcified tissues such as bone or tooth, they are of little relevance in soft tissue, which can 
barely support them at ultrasonic frequencies. 



Longitudinal 
waves are much 
more important 
than shear waves 
in soft tissues 
at diagnostic 
frequencies 



The longitudinal wave is therefore of primary importance for medical applications of 
ultrasound. In a longitudinal wave, individual molecules or particles in the medium 
oscillate sinusoidally about a fi >ed location, moving forward and backward along 
the direction of propagation of the wave energy (Figure 2.1). As the particles move 
forward they become closer to those ahead, so increasing both the local density and 
the local pressure in the medium. Following their maximum forward displacement, 
the particles return towards and beyond their equilibrium location, resulting in a 
slight density reduction, and a reduction in local pressure. The difference between the 
ambient pressure (approximately atmospheric pressure) and the local pressure as the 
wave passes is called the "acoustic pressure". This may be a compression (pressure 
above ambient) or a rarefaction (pressure below ambient). The greatest value of the 
acoustic pressure is of considerable importance when discussing aspects of safety 
concerning mechanical hazard. In particular, the "peak rarefaction pressure" is 
strongly related to cavitation events (see later). In diagnostic scanners these acoustic 
pressures can reach more than 2 MPa at the transducer face, or about 20 atmospheres. 
Referring to the rarefaction pressure, this means that the tissue is being pulled apart 
with a strength equal and opposite to about 20 atmospheres compression. The reason 
that it does not usually rupture is twofold. First, tissue, like water, can withstand this 
stress under many conditions. Second, the stress lasts for a very short time: at 1MHz 
the rarefaction lasts only 0.5 ps, and this period becomes progressively shorter as the 
frequency increases. 

The distance between one compression (or rarefaction) and its immediate neighbour defi 
nes the wavelength, A (Figure 2.1). At any particular frequency,/, the wavelength, A, can 
be calculated from a knowledge of the velocity c (see below), using the expression A = 
c //. At 1 MHz the wavelength in soft tissues is typically between 1.5 mm and 1.6 mm, 



The ultrasonic 
wave consists 
of compressions 
and rarefactions 



Adjacent 
compressions 
are separated by 
one wavelength, 
typically 
0.1-1mm in 
soft tissues 
at common 
diagnostic 
frequencies 



2 The propagation of ultrasound through tissue 



wavelength, A 



• • • 

• • • 



• • «••• • • 

• • «••• • • 

• • m» • • • 

• • »• • • • 

• • • • • 

• • •••• • • 

• • at* • • • 



• • •••• • • • 

• • •••• • • • 



• • •••• • • • 



«••••• 

•••••• 

«••••* 

«••••• 

•»•••• 

•••••• 

«••••• 



• • • •••• • • • 



• • • •••• • • • 



• • •••••• • • 



••• • ••• 

••• • ••• 

• •• •••• 

• •• •••• 

••• • ••• 

• •• •••• 

••• • 



• •••••• • • 



• M* • • • 
••••• • • 



• ••••• 

• •••• • 



• • ••• • • • 



Direction of wave propagation 



Particle movement 
(exaggerated) 



Figure 2.1. A diagram representing the progression of a longitudinal compressional wave 
moving forward by about half its wavelength. The time delay between each wave and the 
one below it is about AY6c 0 , where c Q is the speed of the wave. The dots represent the particles, 
which do not progress with the wave, but oscillate about an undisturbed position. 



whereas at the same frequency the wavelength in bone is between 3 mm and 4 mm, 
because the wave travels about twice as fast in bone as in soft tissue (Table 2.1). 



Standing waves 
are rare in vivo 



Under very specific circumstances a standing wave can also be generated. This occurs 
when part of the energy in a longitudinal compressional wave is reflected back and 
interacts with the incoming wave, forming an interference pattern. Although such an 
arrangement can be generated in the laboratory, it is rare for conditions that may give 
rise to standing waves to occur in an ultrasonic field within the body. Moreover, for 
pulsed ultrasound, interference only occurs transiently, and very close to the reflecting 
surface. 



Wave speed 
depends 
on density 
and elastic 
properties 



2.2.1 Wave propagation speed 

The speed at which an ultrasonic wave propagates is controlled by the mechanical 
properties of the medium. For liquids and soft tissues the speed of the wave, c Q/ depends 
on the compressibility and the undisturbed density p . Solids support both longitudinal 
waves and shear waves, whose speeds depend on the elastic moduli of the solid. However, 
simple equations are difficult to apply directly to biological solids, including bone. This 
is partly because the mechanical properties of some tissues depend on direction, and 

6 



The propagation of ultrasound through tissue 2 



Table 2.1. Representative values for some acoustic properties of tissues at body temperature. 
Note that these are representative values only, and there are very wide variations of tissue 
properties for bone and soft tissues: Blood and amniotic fluid are better characterized. Values 
taken from Duck (1990), ICRU (1998) and Verma etal. (2005). 



Cortical 
bone 



Non-fatty 
tissue 



fat 



Blood 



Amniotic 
fluid 



Propagation speed (m s _1 ) 3635 



Characteristic acoustic 
impedance (10* kg rrr 2 s -1 ) 

Attenuation coefficient at 
1 MHz (dB cm 1 ) 

Attenuation coefficient 
frequency dependence 

Non-linearity coefficient, 
B/A 



6.98 
20 
n/a 
n/a 



1575 

1.66 

0.6 

1.2 

7.0 



1465 



1.44 



1.0 



1.0 



10.0 



1584 



1.68 



0.15 



1.2 



6.1 



1534 



1.54 



0.005 



1.6 



n/a 



consequently so do their ultrasonic properties. This dependence on direction is termed 
anisotropy. 



Values for the wave speed of ultrasound through selected tissues are given in Table 2.1. 
This table gives representative estimates of the speed with which ultrasound propagates 
in the range from 1 MHz to 10 MHz, at body temperature, in normal adult human tissues. 
Tissues from a particular organ, for example the liver, have a range of properties that 
may depend on age, sex, disease state, perfusion and even dietary habits. An increase in 
either water or fat content leads to a decrease in wave speed. Both fatty breast and fatty 
liver tissue have a lower wave speed than comparable normal tissue. Foetal tissues also 
have slightly lower speed than comparable adult tissue, but this is because of their higher 
water content. The presence of collagen, particularly in tendon, skin and arterial wall, 
gives rise to slightly higher speeds than in other soft tissues. 



Speed through 
tissue depends 
on fat, 
collagen and 
water content 



2.2.2 Specific acoustic impedance and interface reflections 

When the particles of the medium move in response to an ultrasonic wave (Figure 2. 
1), there is a particle velocity associated with this movement. (This is quite distinct 
from the speed with which the wave travels.) Oscillations of particle velocity, v, 
and acoustic pressure, p, in a plane progressive wave are in phase: that is, the particles 
move fastest when the acoustic pressure is greatest, p and v are also proportional, 
and the constant of proportionality p/v is called the specific acoustic impedance, Z. 
A simple analysis shows that the acoustic impedance is equal to p 0 c 0 . Knowledge of 
the acoustic impedance of a particular tissue is not, of itself, of great importance. The 
signifi cance of this quantity is demonstrated only when considering the reflection 
and transmission of an ultrasonic wave as it passes across a boundary between two 
materials with different Z, or when small-scale variations in Z result in scattering. 
Acoustic impedance differs litt h between diff eent soft tissues, and between soft tissues 
and water. The greatest diff erences occur at the interface between soft tissue 

7 



Changes in 
specific acoustic 
impedance 
control 
transmission 
and reflection 
at interfaces 



2 The propagation of ultrasound through tissue 



and bone where about one-half of the incident intensity is reflected, and at the interface 
between soft tissue and gas, which reflects almost all the incident wave. This second 
example is also interesting in that it is a so-called "pressure release interface" which 
causes the pressure wave to change phase. The compression in the wave is reflected as 
a rarefaction, and vice versa. The reflection process does not depend on the frequency 
of the wave, the same fraction being reflected from a plane soft-tissue/bone interface 
at 10 MHz as at 1 MHz. 



2.2.3 Attenuation, absorption and scatter of ultrasound by tissue 



Attenuation is 
described as 
an exponential 
loss of pressure 
amplitude with 
distance 



Thus far in the discussion, no mention has been made of energy loss in the tissue through 
which the ultrasonic wave passes. This energy loss, or attenuation, gives rise to energy 
deposition in body tissues. The attenuation of a plane sound wave at a single frequency 
is described by the expression 

p = p 0 e-^ (2.1) 

where the initial acoustic pressure amplitude p 0 has decreased to p x after a travelling 
a distance x (see Figure 2.2). a is the amplitude attenuation coefficient, with units of 
neper per centimetre, Np cm 4 . The relative reduction in amplitude or intensity is often 
expressed on a decibel scale, when the value is 8.68a dB cm 4 . 



Attenuation 
coefficient of 
tissue depends 
linearly on 
frequency, 
approximately 



The attenuation depends on the frequency of the wave. It is greater at higher frequencies. 
For soft tissues the dependence on frequency is approximately linear. It is common 
therefore to give values of the attenuation coefficient for tissue in units of decibel per 
centimetre per megahertz, dB cm 4 MHz 4 . 

Both absorption and scattering contribute to the reduction in acoustic pressure amplitude 
when an ultrasonic wave propagates through tissue. Therefore the total attenuation 




0 20 40 60 80 100 120 140 

Distance, mm 



Figure 2.2. A diagram showing the alteration in amplitude with depth of an ultrasound pulse 
propagating into tissue. This example is for a 3 MHz beam, focused at 70mm, propagating 
through tissue with an attenuation coefficient of 0.5dBcm 4 MHz 4 . 

8 



The propagation of ultrasound through tissue 2 



coefficient a can be expressed as (a a + a), where a a is the absorption coefficient and a s is 
the scattering coefficient. For soft tissues, attenuation is strongly dominated by absorption 
in the low-megahertz range, with scatter losses contributing no more than 10% to the total 
attenuation (Duck, 1990). For calculations involving energy loss the appropriate property 
is the attenuation coefficient for intensity, la. 

The processes by which ultrasonic energy is absorbed by tissues are complex, and not 
fully understood. The frequency dependence diff es from that of a simple liquid like 
water, for which att svuation over this frequency range depends on the square of the 
frequency. Representative values for some tissues are included in Table 2.1, which gives 
both the attenuation coefficient at 1 MHz and its frequency dependence. As a rule of thumb 
the average attenuation coefficient in soft tissue at any frequency is often taken as being 
0.5 dB cm 1 MHz 1 . The fraction of the input energy that is deposited in soft tissue, up to 
specified depths and for beams at 2 MHz, 3 MHz, 5 MHz and 10 MHz is shown in Figure 2.3. 



Both absorption 
and scatter 
contribute to 
attenuation: 
in soft tissue, 
absorption 
dominates 

For most 
diagnostic 
beams, 90% 
of the power 
is deposited 
within the first 
5 cm of tissue 



The scattering of sound from tissue is anisotropic (depends on direction) and arises from 
small-scale variations in density and/or bulk compressibility, and hence in sound velocity. 
In the low-megahertz range there is strong coherent (i.e. in phase) forward scatter with 
generally weak scattering in all other directions. Only the very low-level backscattered 
component contributes to pulse-echo imaging, and this constitutes a vanishingly small 
fraction of the incident energy. The integrated backscattered energy from soft tissue may 
be as low as 50 dB below (that is, 0.00001 of) the incident energy, which implies that 
essentially all of the energy entering the body is deposited in the tissue. 



Essentially all the 
acoustic power 
incident entering 
through the 
skin surface is 
absorbed in the 
body tissues 




■2 MHz 
•3 MHz 
■5 MHz 
■10 MHz 



4 6 8 10 

Depth into tissue, cm 



12 



14 



Figure 2.3. The fraction of the acoustic power leaving the transducer which is deposited in 
soft tissue up to a particular depth, depending on frequency. An absorption coefficient of 
O.BdBcrrr 1 MHz 1 has been assumed. 



9 



2 The propagation of ultrasound through tissue 



Bone attenuates 
much more than 
soft tissue 



Attenuation in bone is much greater than in soft tissue. Attenuation coefficients in the 
range 10-20 dB cm 4 have been reported at 1 MHz for cortical and skull bone. Attenuation 
in trabecular bone is highly variable, probably due to the contribution from scatter. 



Diagnostic 
pulses are 
typically 
shorter than 
1us and 
contain a 
spectrum of 
frequencies 



2.2.4 Beam structure and frequency content 

In practice, a number of other characteristics of beams of sound are significant for 
the complete description of the transmission of ultrasound through tissues. The structure 
of a beam of ultrasound close to its source can be highly complex (Humphrey and Duck, 
1998). Of particular practical interest are the beams from the pulsed transducers that are 
widely used in medical diagnostic applications. Such sources emit very short pulses, 
being typically only two or three cycles, about 0.5 ps, in duration. The energy in these 
pulses of ultrasound is contained in a band of frequencies extending both above and 
below the resonant frequency of the ultrasound transducer that forms the source. 



Focusing 
increases the 
intensity by up 
to 50 times, 
excluding 
attenuation 
effects 



Diagnostic beams are also focused. This is done to reduce the beam width in order to 
improve imaging resolution. Focussing has the additional effect of increasing the acoustic 
pressure and intensity (see below) in the focal zone. The degree of focussing is weak, 
however, giving an increase in pressure amplitude of no more than about a factor 7, 
equivalent to a gain in intensity of about 50. In tissue, this increase is reduced because of 
attenuation of the tissue lying between the transducer and the focus. 



Acoustic power 
is a measure 
of the rate of 
energy flow 



2.2.5 Acoustic power and intensity 

The total acoustic power emitted by the transducer is of central importance when 
considering its safe use. Acoustic power is a measurement of the rate at which energy is 
emitted by the transducer measured in watts: that is, joules per second. Acoustic powers in 
diagnostic beams vary from less than 1 mW to several hundred milliwatts. All this power 
is absorbed by the tissue, and, as a result, the temperature of the tissue is raised slightly. 
Although the power is delivered in very short pulses, it is more relevant to heating to 
average out the effects and to consider only the average power over many seconds. 



Maps of acoustic 
intensity describe 
the spatial 
distribution of 
power 



Whilst acoustic power is important, it is also relevant to describe how that power is 
distributed throughout the beam and across a scanning plane, so that local "hot-spots" 
may be quantified. This variation in "brightness" is measured as acoustic intensity, which 
is obtained by averaging the power over an area. The practical unit of measurement is 
milliwatt per square centimetre, mWcirr 2 . The area may cover the whole beam, or a 
very local part of the beam. A commonly quoted intensity is the "spatial-peak temporal- 
average intensity, I sts ", which is the greatest intensity in the beam, where the beam is 
"brightest" . For an unscanned beam, such as that used for pulsed Doppler or M-mode, this 
will be in the focal zone: for a scanned beam, it may occur much closer to the transducer, 
particularly for sector scan formats. 



Acoustic power and spatial-peak time-average intensity only give information about 
energy deposition when averaged over extended periods of time. Other acoustic quantities 
are used when it is necessary to describe the magnitude of the pulse itself; for example, 

10 



The propagation of ultrasound through tissue 2 



when considering mechanical effects which might result from the interaction of a single 
pulse with tissue, rather than a series of pulses. The most fundamental of these is the 
peak rarefaction pressure, p r . The two other quantities, which are also used to describe 
the magnitude of the pulse, are the mechanical index, which is calculated directly from 
the peak rarefaction pressure (see Chapter 10), and the pulse-average intensity which 
describes the "brightness" of each pulse. 



Rarefaction 
pressure, 
mechanical index 
and pulse-average 
intensity all 
describe the size 
of the ultrasound 
pulse itself 



2.2.6 Estimates of in situ exposure 

It is not generally possible to measure the acoustic field within the body directly. This 
difficulty has meant that alternative methods have been developed to give estimates 
of acoustic quantities such as power, acoustic pressure and intensity within the tissue 
during scanning, so-called "estimated in situ exposure". Ideally, a numerical model 
would be used to predict pulse wave propagation through body tissues, taking account 
of all absorption, scattering, refraction and non-linear processes, and recognizing that the 
body tissues form a three-dimensional distribution of varying acoustic properties. The 
extreme complexity of this approach has led to a practical simplification, which is used at 
present whenever "estimated in situ exposure" is required. 



Very simple 
models are 
generally used to 
estimate in situ 
exposure 



All calculations are based upon measurements of the acoustic pressure in water. The tissue 
is modelled with uniform, homogeneous attenuating properties, with an attenuation 
coefficient of 0.3 dB cm 1 MHz ~\ The selection of this value for attenuation coefficient, 
which is lower than the average for soft tissues alone (see Table 2.1), is justified by the 
view that it safely takes account of propagation through both soft tissue (with a slightly 
higher loss) and fluids (with lower loss). On average this method should overestimate 
the local exposure. Whilst this may be generally true, it must also be emphasized that 
in situ exposures estimated using this very simple model can only be taken as gross 
approximations to actual exposures. 



0.3dBcm- 1 
MHz- 1 allows a 
safety margin 
for estimated in 
situ exposure for 
many situations 



2.3 Non-linear propagation effects 

Thus far the discussion has assumed that the ultrasonic wave is governed by linear laws of 
acoustic propagation. This may be a poor approximation to what actually happens when 
ultrasonic pulses travel through tissue. So-called "finite-amplitude" effects occur, the 
terminology coming from the need to describe theoretically waves apart from those with 
vanishingly small amplitudes. These effects are of practical importance when considering 
exposure measurement, and the biophysical effe cts of ultrasound (Duck, 2002) . An initially 
sinusoidal pressure wave of finite amplitude does not retain its sinusoidal waveform as 
it propagates. The compressions in the wave travel forward faster than the associated 
rarefactions partly because the speed of sound depends on density. This results in a 
distortion of the wave, in which the compressions catch up on the preceding rarefactions, 
ultimately forming a pressure discontinuity or shock. A comparison between the pulse- 
pressure waveform at two distances from a transducer is shown in Figure 2.4. This shows 
the distortion in wave shape, which has been caused by several centimetres travel through 
water, with its accompanying acoustic shock separating the highest amplitude rarefaction 
and compression. The amount of non-linear distortion increases with several factors: the 

11 



Non-linear 
propagation 
causes waveform 
distortion and 
acoustic shock 
formation 



2 The propagation of ultrasound through tissue 



0.06 



0.04 



0.02 



3 

C/> 
HI 

Q. 

U 



0.02 



-0.04 



-0.06 

0 0.5 1 1.5 

Time, microseconds , , 

(a) 



4 



3 

ra 



D- 




-2 ' — 

0 0.5 1 1.5 2 

Time, microseconds ,ys 



Figure 2.4. Two pressure pulses measured in water at the focus of the same 3.5 MHz diagnostic 
transducer, (a) one at low amplitude and (b) the other at high amplitude. The high-amplitude 
pulse shows strong waveform distortion and acoustic shock (an abrupt change from rarefaction 
to compression). 



12 



The propagation of ultrasound through tissue 2 



frequency and amplitude of the wave; the non-linear coefficient of the medium; and the 
distance travelled by the wave. 



As a result of the distortion caused by the non-linear propagation of the wave, its frequency 
content is altered and energy passes from the fundamental frequency into harmonics 
(overtones). The propagation of such shocked waves is associated with additional 
energy absorption, which enhances, sometimes significantly, the propagation losses and 
deposition of energy. Eventually the phenomenon of acoustic saturation occurs. This 
describes the condition where, as the wave amplitude at the transducer is increased, 
none of this additional wave energy arrives at some distance away from the transducer, 
because all additional acoustic energy leaving the transducer is lost through the process 
of excess energy absorption. In practice, the generation of acoustic shocks is common 
when ultrasonic pulses generated by medical imaging systems propagate through water. 
It is predicted that severe waveform distortion and perhaps full shock generation may 
also occur within the fluid spaces in vivo, because of their low attenuation. Examples 
include propagation within urine in the bladder or in the amniotic fluid within a pregnant 
uterus. Propagation through soft tissue inhibits the formation of high levels of harmonic 
because of greater absorption losses. 



Distorted 
waves are rich 
in harmonics, 
resulting in 
increased 
attenuation 



Non-linear effects are significant in discussions of ultrasound safety for two main reasons. 
First, all estimates of acoustic exposure within the body are based on measurements in 
water, in which non-linear effects are strong, and no correction is applied when estimating 
in situ exposure. It has been predicted that acoustic saturation can limit the effectiveness of 
the present Food and Drug Administration limits for the control of ultrasound exposure 
(see Chapter 10), particularly for longer focal depths and higher frequencies (Duck, 1999). 
The second reason is that harmonics can enhance the deposition of energy in tissue, which 
may in turn increase warming and radiation forces. 



In non-linear 
beams in situ 
exposures can be 
underestimated 
and bio-effects 
may be 
accentuated 



2.4 Mechanisms for effects on tissue 

The preceding sections have presented in outline the main important processes that occur 
during the propagation of an ultrasonic wave through tissue. As a result of a variety of 
absorption processes, energy is deposited in the tissue. The response of the tissue will 
depend in part on the mechanism for this deposition, and thus on one of several alternative 
properties of the beam. It is conventional to consider two broad categories: thermal effects 
and mechanical effects. Broadly, mechanical effects can best be predicted from knowledge 
of individual pulses, whilst thermal effects can best be predicted from knowledge of 
energy flow over an extended time period. In addition, as will be detailed below, the 
tissue response is modified considerably by the presence of bone, gas and fluid spaces. 



The two main 
bio-effects 
mechanisms 
are heating 
and mechanical 
processes 



2.4.1 Heating 

Acoustic energy may convert to heat, transferred into the tissue by a variety of absorption 
processes. The rate per unit volume at which heat is produced, dQ/dt, is equal to 2a I, 
where a a is the amplitude absorption coefficient (which increases with frequency) and I is 
the intensity of the wave. The initial rate of temperature rise is equal to 2a 1/ 'C where C is 

13 



2 The propagation of ultrasound through tissue 



Tissue warming 
depends on 
acoustic intensity 
and beam size, 
and on tissue 
absorption 
coefficient, perfusion 
and thermal 
properties 



the heat capacity of the medium. Subsequent heating depends on the width of the beam. 
Broader beams can cause higher temperatures for a given peak intensity than do narrow, 
more highly focused beams. The steady-state temperature also depends on the thermal 
conductivity of the tissue and on the effects of blood perfusion. An "effective thermal 
conductivity" is commonly used in calculations to allow for convective heat loss due to 
blood flow. However, perfusion becomes important only in the wider parts of the beam, 
away from the focal zone. 



Primary bone 
heating is 
markedly higher 
than soft tissue 
warming. Tissue 
adjacent to bone 
will experience 
secondary 
warming 



Tissues with higher absorption coefficients can get warmer than those with less 
absorption. So, the surfaces of calcified bone absorb energy strongly, and heat more than 
soft tissues. Transmission into the bone, and hence its increase in temperature, may be 
reduced for angles of incidence other than those near normal. Foetal bones absorb energy 
more strongly than the surrounding foetal soft tissue, and this difference becomes greater 
as the foetal bones calcify. A 30-fold increase in absorption coefficient has been reported 
as the foetal bone matures (Drewniak et al., 1989). Adjacent soft tissues can experience 
secondary heating from thermal conduction into the tissue from the bone. 



Acoustic 
cavitation occurs 

when bubbles 
are driven by an 

ultrasonic field 



2.4.2 Mechanical effects: cavitation and radiation pressure 

When a gas bubble in a liquid experiences the variations in pressure of an acoustic wave 
its size is driven to change, expanding during the period of decreased pressure and 
contracting during the compression half-cycle of the wave. This behaviour is termed 
acoustic cavitation. For low values of peak acoustic pressure, oscillations in bubble 
radius largely follow variations in pressure. As the peak acoustic pressure increases, the 
bubble becomes unstable as it contracts, collapsing catastrophically under the inertia of 
the surrounding liquid. Such cavitation is therefore termed "inertial" to distinguish it 
from stable or non-inertial cavitation. The term acoustic cavitation is also used to refer 
to the creation of bubbles in a liquid by an acoustic field at nucleation sites, such as 
microscopic impurities, surface roughness on the container or even small-scale local 
density variations. 



Bio-effects of 
acoustic cavitation 
arise from shear 
forces, and free- 
radical formation 



Gas in lung, 
intestine and 

contrast materials 
increases the 
likelihood of 
mechanical 

damage to tissue 



Complex mechanical forces are exerted on the surrounding fluid, on any surface 
adjacent to the bubble, and between one bubble and its neighbours. Biologically, 
probably the most important of these are the shear forces exerted at the bubble 
surface. Mechanical forces of this sort are associated with both non-inertial and 
inertial cavitation, although clearly they are significantly higher in the latter case. 
Chemical action is also possible. The adiabatic conditions associated with extremely 
rapid bubble compression during inertial cavitation result in very high instantaneous 
temperatures within the bubble. These can result in the creation of highly reactive 
free-radical chemical species. 

It is highly improbable that either form of cavitation can be generated at diagnostic 
levels within soft tissues or fluids in the body, in the absence of gas-filled ultrasound 
contrast agents. However, there are two conditions when the presence of gas may result 
in mechanical trauma to adjacent soft tissue, caused by a cavitation-like process: at the 
surface of the lung, and in the intestine. 

14 



The propagation of ultrasound through tissue 2 



Finally, tissues may experience a range of other forces from the passage of an ultrasonic 
wave (see Chapter 6). In particular, a radiation stress is exerted within tissues and fluids as 
the pulse propagates, and also at interfaces where there is a change of acoustic impedance. 
When exerted within a liquid this force causes acoustic streaming, and the fluid moves in 
the direction of the pulse propagation. This radiation stress is of much lower magnitude 
than that associated with bubble activity, but exists universally and does not require the 
presence of gas bodies. 



Low- 1 eve I 
radiation 
stress always 
accompanies 
ultrasound wave 
propagation 



2.5 The passage of an ultrasonic pulse through tissue 



Based on the preceding discussion, and at the risk of some minor repetition, we are 
now in a position to follow what happens when a real ultrasonic transducer generates a 
series of acoustic pulses, which then propagate through tissue. The pulses are generated 
by a broadband piezoelectric transducer. Such transducers are inherently poor in their 
efficiency of transferring electrical energy to acoustic energy, and as a result heat is 
dissipated in the transducer: it warms up. It is probable that the greatest tissue heating 
during diagnostic ultrasound arises from this cause (Calvert et ah, 2007), and it should 
be considered seriously when thermally sensitive tissues lie close to the transducer, as in 
ophthalmic scanning. 



Pulsed ultrasound 
transducers 
generate heat 



The penetration of the pulse into the tissue depends on the effectiveness of the acoustic 
coupling to the tissue. For skin-coupling the attenuation coefficient of the dermal and 
sub-dermal layers may also have a strong effect, since it may be high depending strongly 
on hydration, and fat and collagen content. The acoustic pulse contains a broad spectrum 
of frequencies centred approximately at the resonant frequency of the piezoelectric 
source. The amplitude and intensity of the wave reduces with distance at a rate of about 
0.5 dB cm 1 MHz -1 ; for a 3.5 MHz wave, the amplitude will be reduced by one-half, and the 
intensity by a factor of four (-6 dB) after travelling about 4 cm, mostly due to viscous and 
relaxation absorption processes. The remaining energy is scattered, effectively spreading 
the beam, and this scattered energy may undergo further scattering interactions. An 
extremely small fraction of the energy returns to the transducer. 



The majority of 
the transducer 
output power is 
absorbed in the 
superficial tissue 
layers 



If there is a repetitive sequence of pulses, as in most diagnostic applications, the tissue 
will be warmed as a result of the absorption of acoustic energy. The temperature rise 
depends on the time-averaged acoustic intensity, the acoustic absorption coefficient, the 
thermal properties of tissue (heat conduction and specific heat), tissue perfusion (blood 
flow), beam size and scanning mode and the period of time the transducer is held in one 
position. The tissue also experiences a small transient force in the direction of propagation 
each time a pulse passes. If the pulse passes through a liquid, it will move in the direction 
of the pulse propagation: a series of pulses will cause acoustic streaming. 



The tissue is 
slightly warmed, 
and slightly 
stressed during 
diagnostic 
scanning 



The pulse spectrum alters as the wave propagates. In soft tissue this alteration is 
dominated by the frequency-dependent attenuation of the tissue. As a result, higher 
frequencies in the pulse spectrum reduce in proportion to those at lower frequencies, 
so lowering the mean frequency in the spectrum of the pulse. For a pulse of very high 
amplitude, finite-amplitude effects also come into play and some energy is passed to 

15 



The pulse 
frequency 
spectrum alters 
as the pulse 
propagates 



2 The propagation of ultrasound through tissue 



higher-frequency harmonics. This latter effect is more pronounced during transmission 
through fluids, however, where it is the dominant mechanism modifying the pulse 
spectrum. 



Bone heats 
preferentially, 
and will warm 
surrounding 
tissues 



As the wave propagates farther into the tissue it may reach a clear acoustic interface 
between media of differing acoustic properties. If the second medium is bone, about 
half the energy in the wave is reflected and half enters the bone. The pattern of 
reflected energy will depend somewhat on the scattering properties of the tissue-to- 
bone boundary, and the subsequent propagation of this scattered wave through soft 
tissue is difficult to predict. Standing waves are very unlikely to form. The remaining 
energy that enters the cortical bone may propagate as longitudinal, shear or surface 
waves, all of which are rapidly absorbed, resulting in a local temperature rise. This 
bone heating causes secondary heating of the surrounding soft tissues by thermal 
conduction. 



Tissues next to 
a gas boundary 
are particularly 
vulnerable to 
mechanical 
damage 



Almost all of the incident wave energy is reflected from any boundary between soft 
tissue and gas. This gas may exist within the alveoli of the lung, within the intestine or 
at the exit site of the beam. Also, gas bubbles may be artificially introduced to act as a 
contrast medium in blood. Such tissue-to-gas interfaces constitute very large alterations 
of acoustic impedance and the resulting pressure wave is, to a first approximation, 
of equal amplitude and opposite phase to that of the incoming wave. Mechanical 
stress experienced by soft tissue at a tissue-to-gas interface can be sufficient to cause 
permanent damage to membranes (causing lysis of erythrocytes in the presence of 
bubbles, for example) or to weak connective tissue structures, especially tissues with 
low shear strength (causing, for example, lung capillary bleeding). Were inertial 
cavitation to occur, extreme conditions of temperature and pressure could be locally 
generated, which in principle could lead to free-radical generation. This has not been 
demonstrated in vivo. Apart from mechanical effects, the interaction between the 
acoustic wave and bubbles can also generate heat locally, because of a general increase 
in absorption coefficient. 



Another interface of interest is that from soft tissue into fluid. Little energy is reflected 
since the acoustic impedance change across the boundary is slight. The wave emerges 
into a space containing, for example, blood, amniotic fluid or urine. Scatter is minimal, 
absorption is low and finite-amplitude distortion processes are not strongly suppressed. 
The wave therefore carries frequency components through the fluid that are substantially 
higher than those generated by the transducer, especially in the focal zone. When this 
pulse reaches a further fluid-to-tissue boundary, much of its high frequency content will 
be deposited in the superficial tissue layers, leading to greater warming and radiation 
stress than from equivalent undistorted pulses. 



Liquids in vivo 
accentuate 
non-linear 
effects 



2.6 Conclusion 

The propagation of ultrasound and the mechanisms of action between ultrasonic waves 
and tissue are now well understood. The generation of this knowledge has been largely 
stimulated by the widespread use of ultrasound in the low-megahertz frequency 

16 



The propagation of ultrasound through tissue 2 



range in diagnostic and therapeutic medicine. Much is still unclear, however, about 
the detailed interaction at a microscopic level of these interactions and mechanisms. 
Furthermore, the thresholds and conditions for cavitation, and the importance of finite- 
amplitude transmission within tissue, and the relevance of radiation stress still require 
clarification. 

References 

Calvert J, Duck F, Clift S, Azaime H. 2007. Surface heating by transvaginal transducers. 
Ultrasound Obstet Gynecol, 29, 427-432. 

Drewniak JL, Carnes KI, Dunn F. 1989. In vivo ultrasonic heating of fetal bone. JAcoust Soc 
Am, 86, 1254-1258. 

Duck FA. 1990. Acoustic properties of tissue at ultrasonic frequencies. In Physical 
Properties of Tissue, a Comprehensive Reference Book. London, UK: Academic Press, 
pp. 73-135. 

Duck FA. 1999. Acoustic saturation and output regulation. Ultrasound Med Biol, 25, 
1009-1018. 

Duck FA. 2002. Nonlinear acoustics in diagnostic ultrasound. Ultrasound Med Biol, 28, 
1-18. 

Humphrey VF, Duck FA. 1998. Ultrasonic fields: structure and prediction. In Ultrasound 
in Medicine, Duck FA, Baker AC, Starritt HC (editors). Bristol, UK: Institute of Physics 
Publishing, pp. 3-22. 

ICRU. 1998. ICRU Report 61: Tissue Substitutes, Phantoms and Computational Modelling 
in Medical Ultrasound. Bethesda, MD: International Commission on Radiation Units and 
Measurements. 

Verma PK, Humphrey VF, Duck FA. 2005. Broadband measurements of the frequency 
dependence of attenuation coefficient and velocity in amniotic fluid, urine and human 
serum albumin solutions. Ultrasound Med Biol, 31, 1375-1381. 



17 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 3 

The acoustic output of diagnostic 
ultrasound scanners 

Adam Shaw 1 and Kevin Martin 2 

^Acoustics and Ionizing Radiation Division, National Physical Laboratory, Teddington, UK 
2 Department of Medical Physics, University Hospitals of Leicester, Leicester, UK 



Summary 

• Four important acoustic output quantities are the peak rarefaction pressure (p r ), the 
spatial-peak temporal-average intensity (I ), the temporal-average acoustic power 
(W) and the temperature of the transducer face (T ,). 

• The measurement of acoustic outputs in clinical environments requires appropriate 
equipment and techniques. 

• In general, I s ^ W and T surf are greatest for spectral Doppler mode and least 
for B-mode. For all three quantities there is considerable variation between 
different transducers and machine models. Values of p r do not vary much between 
modes. 

• Surveys since 1991 demonstrate that p r values have increased steadily. I ^ values 
in B-mode have shown the greatest increases and now overlap the range of pulsed 
Doppler values. 

• Maximum mechanical index values declared by manufacturers are biased towards 
the Food and Drug Administration (FDA) maximum permitted level. Manufacturer 
declared values of thermal index are on average much lower than the FDA normal 
maximum level, but still significant in relation to acoustic safety in obstetric and 
neonatal scanning. 

In the previous chapter, some of the parameters that may be used to characterize the 
beams and pulses from diagnostic ultrasound systems have been described. It was shown 
that these parameters could be used to assess the likelihood of tissue heating or cavitation 
during exposure. The aim of this chapter is to explain how relevant acoustic parameters 
can be measured for diagnostic systems and how these parameters are affected by user 
controls. Values of acoustic parameters and their trends for modern diagnostic systems 
are also reviewed. 



18 



The acoustic output of diagnostic ultrasound scanners 3 



3.1 Acoustic output parameters 

3.1.1 Acoustic pressure 

A point within the ultrasound beam experiences successive cycles of compression and 
rarefaction during the passage of an ultrasound pulse (see Figure 3.1a). 



The magnitude of the pressure changes is characterized by the peak compression and 
rarefaction pressures, which are the greatest values during the pulse. The peak rarefaction 
pressure p r can be used in assessing the risk of occurrence of cavitation or other gas- 
body activation events. The peak rarefaction pressure changes with position in the beam 
and is greatest in the focal region. Acoustic pressure is normally measured in water 
using a hydrophone (see later). 



The peak 
rarefaction 
pressure in 
an ultrasound 
beam is used 
to assess the 
risk of cavitation 



3.1.2 Acoustic power 

Each pulse transmitted into the tissue medium carries acoustic energy [measured in joules 
(J)] which is gradually absorbed and deposited in the tissue. The rate at which energy is 
transmitted into tissue from the transducer by this means is the total acoustic power W, 



Pressure 



(a) 




Peak compression pressure (p c ) 



Static pressure 



Absorption 
of acoustic 
power in 
tissues causes 
tissue heating 
and radiation 
stress 



Peak rarefaction pressure (p r ) 



Temporal peak intensity (/ tp ) 



(b) 




Pulse-average intensity (/ pa ) 



Time 



Figure 3.1. (a) The peak compression and rarefaction pressures are the maximum and minimum 
values of pressure in the medium during the passage of an ultrasound pulse, (b) The intensity 
is related to the pressure squared and is always positive. The temporal-peak intensity is the 
maximum value during the pulse. The pulse-average intensity is the average value over the 
duration of the pulse. 

19 



3 The acoustic output of diagnostic ultrasound scanners 



Intensity is related 
to the square of 
acoustic pressure 
and is always 
positive 



measured in watts (W) (1 W = 1 J s _1 ). This is the average power over many transmit pulse 
cycles. A medium which absorbs the acoustic power from the transducer of course heats 
up, but it is also subject to a radiation force, which is proportional to the absorbed power. 
This force results in stresses within the tissue [an effect which is now being exploited in 
acoustic radiation force imaging] and in acoustic streaming within fluid-filled regions 
such as the bladder or cysts. 

3.1.3 Intensity 

The intensity in the ultrasound beam is a measure of the flow of acoustic power through a 
given cross-sectional area and is measured in W m 2 or mW cirr 2 (see Figure 3.2). In plane 
waves, intensity is related to the square of acoustic pressure by the equation: 



where Z is the acoustic impedance of the medium (see Chapter 2). Hence, intensity values 
can be derived from measurements of acoustic pressure. 

During the passage of an ultrasound pulse, the pressure and hence the intensity, vary 
with time. Figure 3.1 shows corresponding pressure and intensity waves during a pulse. 
Note that because intensity is related to the square of pressure, its value is always positive. 
The peak value of intensity during the pulse is called the temporal-peak intensity I . 
An alternative and more widely used measure of intensity during the pulse is the 
pulse-average intensity I . The pulse-average intensity is more useful as it is more 
immune to changes in the shape of the pulse than the temporal-peak intensity. 



I = p 



< 2 /Z 



(3.1) 



w 




lm 



Figure 3.2. Intensity is the power W in watts flowing through unit area, e.g. IWr 2 . 

20 



The acoustic output of diagnostic ultrasound scanners 3 



Intensity 



Pulse repetition period 




Temporal- average intensity (7 ta ) 



z 




Time 



Figure 3.3. The intensity waveform is repeated with every pulse-echo cycle. The temporal- 
average intensity is the average value over a complete pulse-echo cycle and is much lower 
than the pulse-average intensity. 

Where the ultrasound beam is stationary, e.g. for a pulsed Doppler beam, the pulse 
waveform is repeated at the pulse repetition frequency. Where the longer term effects 
of exposure to the beam are of interest, e.g. in assessing potential for tissue heating, it 
is useful to measure the temporal-average intensity (I ). This is the value of intensity 
averaged from the start of one pulse to the start of the next (or other similar point). This 
value is much lower than I a as it includes the long "OFF" period between pulses (Figure 
3.3). For scanned modes, where the beam is swept through the region of interest, the 
point of interest in the tissue may be exposed only once in each scan of the beam and the 
average must be taken over a complete scan repetition period. 

As described in Chapter 2, the intensity varies with position in the beam as well as with 
time. Hence it is possible to specify intensity at a particular location in the beam, such 
as at the point where it is maximum. This is the spatial-peak value. Alternatively, it is 
possible to calculate a value averaged over the beam cross-sectional area, known as the 
spatial average value (Figure 3.4). The following are the most commonly quoted intensity 
parameters. 

• 7 sppa (Spatial-peak pulse-average intensity): The pulse average-intensity measured at 
the location where it is maximum. 

• I s u (Spatial-peak temporal-average intensity): The temporal-average intensity 
measured at the location where it is maximum. 

• I (Spatial-average temporal-average intensity): The temporal-average intensity 
averaged across the beam cross section (at a particular range from the transducer). 

3.1.4 Free-field and derated values 

When acoustic pressure or intensity is measured using a hydrophone, the measurement 
is normally made in water, which has almost no attenuation. These are normally called 
free-field quantities. To estimate pressure values that might exist in soft tissue in the 
same ultrasound beam, the measured pressure values are "derated", by an amount that 
depends on the attenuation of the tissue. Most soft tissues have an attenuation of between 
0.5 and 1.0 dB cm -1 MHz 1 . When calculating the safety indices (next section), a lower 
attenuation of 0.3 dB cm 1 MHz 1 is assumed (in case some of the path is fluid-filled) and 
the derated value of peak rarefaction pressure is denoted as p iQ3 . That is, the value p r 



Intensity varies 
with time 
and position 
in the beam. 
Temporal- 
average 
intensity is 
much lower 
than that 
during the 
pulse 



Pressure and 
intensity are 
normally measured 
in water using 
a hydrophone. 
These "free-field" 
values may then be 
derated to estimate 
the values that 
would be expected 
in tissues, assuming 
an attenuation of 
O.SdBcrrr 1 MHz- 1 



21 



3 The acoustic output of diagnostic ultrasound scanners 



Intensity 




Spatial-peak intensity (7 sp ) 




Spatial average intensity (Y sa ) 



Figure 3.4. The values of the various intensity parameters change with position in the beam 
also. The highest value in the beam is the spatial-peak intensity. The average value over the 
area of the beam is the spatial average intensity. 

measured in water, is reduced by 0.3/zdB to estimate the value p r03 that would exist in 
tissue, where /is the frequency of the pulse (in MHz) and z is the range (in cm) from the 
transducer at which the measurement was made. This is a reasonable approximation to 
a "worst-case" safety index, but it does not usually give the best estimate of the field in 
real tissue. Nevertheless, in most cases where "derated" values are reported, the derating 
factor used is 0.3 dB cnr 1 MHz 1 . 

The same process may be applied to values of intensity such as I u or I . Such derated 

JT J r r J spta sppa 

values of intensity are used by the Food and Drug Administration (FDA) in the USA to 
regulate output levels from diagnostic ultrasound systems (see Chapter 10). 



Safety indices Ml 
and Tl are used 
to indicate the 
risk of cavitation 
and the probable 
temperature rise 
in tissue exposed 
to an ultrasound 
beam 



3.1.5 Safety indices 

The pressure, intensity and power parameters detailed above describe the acoustic field 
in water (free-field) or in tissue (derated) and are related to the field that the patient 
is exposed to during diagnosis. Such parameters were widely used to monitor the 
acoustic outputs of early ultrasound systems. On their own however, they were not 
good indicators of the risk of adverse effects. Current standards and regulations refer 
to parameters that are intended to relate more directly to cavitation and tissue heating: 
these are the mechanical index (MI) and the thermal index (TI) (IEC62359, 2010; see 
also Chapter 10). MI is intended to indicate the probability of occurrence of inertial 
cavitation, while TI is an indicator of the likely maximum temperature rise in tissues 
exposed to the ultrasound field. Although these indices can provide useful information 
to the user, they are not perfect and are based on a set of very specific assumptions. 
A particular criticism of TI is that it ignores the self-heating of the transducer, and 
so greatly underestimates the temperature rise within about 5 mm of the transducer. 
For some applications this region may contain sensitive tissue (see later section on 
Measurement of Temperature). 



3.1.5.1 Mechanical index 

MI is defined by: 



MI 



Pr,0 



V7 



(3.2) 



22 



The acoustic output of diagnostic ultrasound scanners 3 



Here, p r03 is the maximum value of derated peak rarefaction pressure (in MPa) in the 
beam. This is measured by recording p r at a range of depths in water and derating the 
values to find the maximum value of p r03 . The centre frequency in the pulse is /MHz. 
The equation for MI is based on a model which assumes the presence of bubble nuclei in 
the tissue. It predicts that inertial cavitation is more likely at higher values of p r03 and at 
lower frequencies. According to the theory, cavitation should not be possible at values 
less than 0.7. 



3.1.5.2 Thermal index 

TI gives an indication of the likely maximum temperature rise in tissue due to absorption 
of ultrasound. 



It is defined by: 



TI = W/W d 



(3.3) 



Here, W is the current acoustic output power from the transducer and W deg is the power 
required to raise the temperature of the tissue by 1 °C. The likely maximum temperature 
rise depends on the type of tissue and on the operating conditions of the ultrasound 
system. As temperature rise in tissue is strongly influenced by the presence of bone, 
3 versions of TI are used to model the anatomical conditions. These are (i) TIS, which 
assumes exposure of uniform soft tissue, (ii) TIB, which assumes that a layer of bone is 
present in or close to the focal region of the beam and (iii) TIC, a model which assumes 
the presence of bone just under the tissue surface. American Institute for Ultrasound 
in Medicine and International Electrotechnical Commission (IEC) standards include 
agreed formulae for calculating the indices for scanned and unscanned beams and for 
beam apertures greater than or less than 1 cm 2 . These standards include a requirement 
that MI or TI is displayed to the user if either value can exceed 1.0 under any operating 
condition. 



Three different 
models are used 
to calculate TI to 
account for the 
presence and 
relative position 
of bone in the 
beam. These are 
TIS, TIB and TIC 



3.1.6 Transducer temperature rise 



In addition to heating due to absorption of ultrasound, the temperature of tissues 
near the transducer is strongly influenced by the temperature of the transducer itself. 
Ultrasound pulses are produced by applying an electrical signal to the transducer. 
Some electrical energy is dissipated in the element, lens and backing material, causing 
transducer heating. Electronic processing of received signals in the transducer head 
may also result in electrical heating. Conduction of heat from the transducer face can 
result in temperature rises of several degrees Celcius in superficial tissues. Maximum 
allowable transducer surface temperatures (T f ) are specified in IEC standards 
(see Chapter 10). These are 50 °C when the transducer is transmitting into air and 
43 °C when transmitting into a suitable phantom. This latter limit implies that skin 
(typically at 33 °C) is permitted to be heated by up to 10 °C. Transducer heating is a 
significant design consideration in complex transducers and in some circumstances 
these temperature limits may effectively restrict the acoustic output that can be 
achieved. 

23 



The temperature 
of the 

transducer face 
can be raised 
due to electrical 
energy losses in 
the transducer, 
resulting in 
heating of 
adjacent tissues. 
Transducer 
temperature 
is limited by 
international 
standards 



3 The acoustic output of diagnostic ultrasound scanners 



3.2 The need for independent measurement 



Independent 
measurements 
are important 
to make sure 
the scanner 
meets necessary 
standards, 
to check 
manufacturers' 
data and to 
support research 



Modern ultrasound scanners have become so complex and have so many different 
output combinations that it is effectively impossible for anyone other than the original 
manufacturer or a very specialized laboratory to attempt a full measurement of the output. 
So, with most modern scanners displaying thermal and mechanical indices (see also 
Chapter 10 on regulations) why would anyone else want to undertake any measurements? 
Of course, not all hospitals can be expected to make complex output measurements but 
there are several important reasons why there must be a capability within any healthcare 
system to make detailed independent measurements. 



The first is a duty of care to those being scanned and the need to ensure that equipment is 
"fit for purpose", meets all necessary standards and is properly maintained. Measurements 
may be needed prior to acceptance on purchase, for routine QA in compliance with local 
or national quality systems, or when a potential fault is reported. Software upgrades 
by field engineers pose a special problem, since the systems are controlled by software 
and the acoustic output may potentially change whenever such an upgrade takes place. 
Report 102 from the Institute of Physics and Engineering in Medicine (IPEM, 2010) deals 
with QA in detail. 

The second reason is to act as a check on the manufacturers. CE marking of equipment in 
Europe is mostly based on self-declaration by the manufacturer, and not on independent 
evaluation. Although the CE marking process in a company is in principle subject to audit, 
this essentially concentrates on management systems, not on the "correctness" of acoustic 
measurements. Apart from measurement problems, record-keeping lapses are always 
a possibility with the ongoing process of hardware and software revisions potentially 
leading to the output of a particular machine being substantially different to that of another 
machine, apparently of the same make and model and revision, or to the manufacturer's 
published value. It is perhaps not surprising, therefore, that manufacturers' reported 
output data has not always been completely reliable (Jago et al, 1995). 



The third reason is to support research, for example into biological response to ultrasound 
or into the use of higher output modes for diagnosis, especially of the foetus, embryo 
or neonate, or of the brain and central nervous system. It is tempting but completely 
wrong, to suppose that the on-screen MI and TI is enough to somehow "characterize" the 
exposure. Anyone thinking about undertaking research must think carefully in advance 
about the output measurements required to support it: ter Haar et al. (2011) have presented 
guidelines for correctly reporting exposure conditions. 



3.3 Measurement methods and equipment 

A detailed description of how to make output measurements is beyond the scope of this 
book but the following sections give a description of the principles, considering pressure, 
intensity, output power and temperature rise. Report 102 from the Institute of Physics and 
Engineering in Medicine (IPEM, 2010) gives practical guidance for which measurements 
are suitable for QA and the maintenance of diagnostic scanners; more general advice 

24 



The acoustic output of diagnostic ultrasound scanners 3 



on sett hg up and using measurement systems can be found in Preston (1991), Lewin 
and Ziskin (1992), and Szabo (2004). The measurement of temperature rise is perhaps the 
most easily accessible measurement, with output power not too far behind. Measuring 
pressure or intensity distributions is a much more specialized task; nevertheless, it is 
logical to deal with this most complex issue first. 

3.3.1 Measurement of pressure and intensity 
3.3.1.1 Pressure 

The fundamental component that allows the acoustic pressure to be monitored and measured 
is the hydrophone. This is essentially a high frequency microphone for use underwater that 
produces a voltage waveform when it is placed in an ultrasonic field. The type preferred for 
use in diagnostic fields is the membrane hydrophone (available from, e.g. Precision Acoustics 
Ltd, Dorchester, UK; ONDA Corporation, Sunnyvale, CA; and Sonora Medical Systems 
Ltd, Longmont, CO, amongst others) which is chosen for its even response over a wide range 
of frequencies. In this type of hydrophone, the pressure-measuring element is the small 
central area of a polymer membrane (polyvinylidene fluoride— PVDF), stretched across a 
circular support ring (Figure 3.5). The diameter of this ring (typically 80 mm) is sufficient to 
accommodate the full B-mode ultrasound field from a medical ultrasound probe. A small 
central region of the film (typically less than 0.5 mm in diameter) is piezoelectrically active: 
it is desirable for this area to be as small as possible, since the voltage produced represents 
the spatial average of the acoustic pressure over this area. 

The sensitivity of a membrane hydrophone increases slowly with frequency up to its 
resonance frequency (approximately 60 MHz for a 16 pm thick membrane). However, a 
more constant frequency response can be obtained by amplifying the hydrophone output 
in a low-noise preamplifier with a frequency response that is deliberately matched to be 
complimentary to that of the hydrophone. Due to non-linear propagation effects which 
can result in the generation of high frequency harmonics in the pulse waveform, the 
international standard IEC62127-1 (2007) recommends that the output of the hydrophone 
and preamplifier should vary by less than ±6dB over a frequency range extending to 
3 octaves above the acoustic working frequency, or to 40 MHz, whichever is the smaller. 



Figure 3.5. Membrane hydrophone (left) and needle hydrophones (right). Photographs 
courtesy of Precision Acoustics Ltd. 



Acoustic 
pressure is 
measured with 
a calibrated 
hydrophone 
connected to an 
oscilloscope 




25 



3 The acoustic output of diagnostic ultrasound scanners 



Good quality modern membrane hydrophones will normally meet this requirement, even 
without a matched amplifier. 



Membrane 
hydrophones 
are more 
accurate 
but probe 
hydrophones 
are cheaper 



Although membrane hydrophones offer the most faithful reproduction of the pressure 
waveform, they are expensive (approximately £8000 in 2011). Another type of hydrophone 
is the probe (or needle) hydrophone in which the active PVDF element is placed on the 
end of a hollow tube (or needle). The needle perturbs the ultrasound field resulting 
in a sensitivity which varies more strongly with frequency, especially below 3 MHz. 
However, they are cheaper to buy (from about £1200 in 2011), and their needle shape 
offers advantages in some situations. Fibre-optic hydrophones are also available in which 
the pressure field modulates the amount of light reflected from the end of an optical 
fibre but these are more normally used in high intensity therapeutic ultrasound (HITU or 
HIFU) fields where piezoelectric hydrophones might get damaged. 



Intensity is 
not measured 
directly. It is 
calculated from 
the pressure 
waveform 



Generally, the hydrophone is mounted in a water tank on a three-axis manipulator that 
allows it to be moved within the transducer field. The signal from the hydrophone is 
monitored on an oscilloscope, from which p r can be measured directly. The use of a digital 
oscilloscope allows other pulse parameters, such as centre frequency, to be calculated 
in real-time if required. The hydrophone calibration should be traceable to national 
standards (for instance at the National Physical Laboratory). Measurement uncertainty 
depends on the particular application but it is typically less than ±10% for acoustic 
pressure parameters, and ±21% for intensity parameters (95% confidence level). 

3.3.1.2 Intensity 

Intensity is a measure of the rate of energy flow through an area. Although there are some 
prototype sensors for measuring intensity based on heating (Wilkens, 2010a,fr; Hodnett 
and Zeqiri, 2009; Zeqiri et al., 2011), in practice intensity is not measured directly but 
is calculated from measurements of pressure using a hydrophone as described in the 
previous section on pressure. The basic assumption that is made in calculating intensity 
is called the "plane-wave assumption" which says that the instantaneous intensity, I(t), is 
related to the instantaneous pressure, p(t) by the relationship: 



Kt) 



v\t) 

pc 



(3.4) 



where p is the density of water and c is the speed of sound in water. Although this 
relationship is not strictly true everywhere, it is a good approximation throughout most 
diagnostic fields and is used in international standards (IEC62127-1, 2007). 



To measure 
i . the 

ta' 

contributions 
from all 
pulses in the 
scanframe must 
be included 



The determination of temporal-average intensity, I , is particularly challenging for 
scanned imaging modes because many separate pulses contribute to the energy flowing 
through a particular point. Modern, deep memory digital oscilloscopes have, however, 
made it much easier than in the past, since it is now possible to capture every pulse in 
the scanframe— even for combined modes — as long as a trigger signal can be obtained 
from the scanner. In the absence of such a dedicated trigger signal, it is possible to trigger 
directly on the hydrophone waveform to capture all pulses that exceed some small 

26 



The acoustic output of diagnostic ultrasound scanners 3 



pressure. Of course this will miss the contribution from the smallest pulses and can also 
capture electrical pickup which is not part of the acoustic signal. An analogue alternative 
is the method developed by Martin (1986) in which the signal from the hydrophone is first 
amplified in a power amplifier and then input to a commercial electrical power sensor to 
measure the time-averaged electrical power generated by the hydrophone without the 
need for a synchronizing trigger: this is proportional to the temporal-average intensity. 
A digital equivalent to this is to continuously digitize the hydrophone signal. With both 
the analogue and digital versions, the electrical noise power should be assessed and, if 
significant, corrected for. 



3.3.1.3 Hydrophone measurement systems 

Turnkey commercial hydrophone measurement systems which integrate a measurement 
tank, positioning system, hydrophone, digital oscilloscope and software are available from 
Precision Acoustics Ltd, Onda Corporation and Sonora. They are not designed to deal with 
scanned modes of operation "out of the box" but additional software capture and analysis 
routines can be written to do this. Although they are sometimes mounted on a very large 
trolley, these systems are essentially more suited to use in a fixed location (Figure 3.6). 



Complete 
hydrophone 
measurement 
systems can 
be bought 



A measurement system which is no longer commercially available but is still being used in 
some centres, and which was designed specifically to deal with scanned operating modes, 
is the NPL Ultrasound Beam Calibrator (Preston; 1988; Shaw and Preston, 1995). This is 
a sophisticated system, based on a linear array of hydrophone elements whose outputs 




Figure 3.6. The UMS3 hydrophone scanning system. Photograph courtesy of Precision Acoustics 
Ltd. 

27 



3 The acoustic output of diagnostic ultrasound scanners 



are sampled in rapid succession to give an effectively real-time profile of the pressure or 
intensity distribution across a beam. The array is formed on a PVDF membrane stretched 
across a support ring, in a similar way to the single element membrane hydrophone 
described previously. Multielement array hydrophones are available from Precision 
Acoustics Ltd. 



Those preferring to build their own measurement system for lower cost could also look 
at the approach taken in Newcastle General Hospital (UK). They designed a lightweight 
and compact system, suitable for transportation by car between base and a hospital 
site, and by a small trolley within a hospital. Since access to most scanning systems is 
limited by the need to cause minimal disruption to the normal clinical workload, it was 
designed to be quick and easy to assemble at the measurement site. It has been used to 
make measurements on over a thousand combinations of probes and modes on a wide 
range of diagnostic ultrasound scanners in the north of England. It uses the analogue 
method developed by Martin (1986) for monitoring 7 fa in real-time. A block diagram of the 
system is shown in Figure 3.7. 



3.3.2 Measurement of output power 



Power is best 
measured using 
a calibrated 
RFB fitted with 
an absorbing 
target 



A radiation force balance (RFB) provides a convenient way of measuring the acoustic 
power from diagnostic equipment in hospital departments. This method makes use of 
the fact that ultrasound exerts a force on a target that is directly proportional to the total 
power absorbed or reflected by the target [guidance can be found at http://www.npl. 
co. uk/acoustics/ultrasound/research/best-practice- guide- to-measurement-of-acoustic- 
output-power-(introduction)]. It is preferable to use a flat absorbing target rather than 
the conical reflecting target which is sometimes seen. The use of a flat target simplifies 
corrections for non-perpendicular incidence (see below), and allows the distance between 
the probe and the target to be reduced, thereby reducing errors due to absorption of high 
frequency harmonics associated with the non-linear propagation of ultrasound waves in 
water. Conical targets should not be used in focused fields (IEC61161, 2006). 



The force on an absorbing target is approximately 68u.gpermW of incident power. 
Since the output power of diagnostic scanners is typically between a few milliwatts and 



Mi cromanipulator 

'A 



Absorber 



Oscilloscope (p r ) 



Power 
amplifier 




Thermocouple 
sensor 



RF Power meter (/ ta ) 



Figure 3.7. Principle features of the portable "hydrophone in a bucket" system developed in 
Newcastle. 



28 



The acoustic output of diagnostic ultrasound scanners 3 



a few hundred milliwatts, a balance resolution of 0.01 mg is required for measurements 
at the lower end of this power range although, with care, a resolution of 0.1 mg is 
adequate for powers above about 20 mW. This high sensitivity means that air currents 
and vibrations transmitted from the surroundings can be a problem. 

The balance calibration should be traceable to national standards (e.g. at the National 
Physical Laboratory), its performance may be checked using a checksource (a transducer 
and drive unit that delivers a beam of known acoustic power) or, in some cases, by 
applying a known weight. The use of a checksource is preferable to weights since it 
will also verify the acoustic performance of the target. If the ultrasonic field is strongly 
convergent (e.g. some strongly focused stationary beams), divergent (e.g. sector scanners 
in scanning mode) or obliquely incident on the target (e.g. angled Doppler or sector scan 
beams), then an appropriate "cos 6" correction factor should be estimated and applied. 
With care, an uncertainty of 10-15% is achievable. 

Commercial RFBs suitable for the diagnostic power range are available from Onda 
Corporation, Ohmic Instruments Co. (Easton, PA) and Precision Acoustics. All use a top- 
loading configuration (see Figure 3.8): in the first two, the target is suspended in a small 
water tank; in the third, the target actually forms the water tank in what is sometimes 
called an "acoustic well" (Sutton et ah, 2003). Some large transducers may not fi tall of 
these RFBs. Especially when measuring powers less than 50 mW, the transducer must 
be held solidly in a stationary clamp, the transducer cable should be supported to stop it 
swinging or moving, and it may be necessary to cover the RFB to protect it from draughts 
(even a cardboard box can be very eff ective). 

It is quite possible to make a RFB. The most important consideration is the need for 
a high quality absorbing target: the best material for this the 2-layer polyurethane 
material called "HAM- A" from Precision Acoustics Ltd (Zeqiri and Bickley, 2000). 

Several centres still use a design of RFB which is no longer available and which is often 
called the "Bath Balance" (Perkins, 1989). This was a development of an earlier design 
by Farmery and Whittingham (1978). This is a closed system in which the transducer 
is placed horizontally against a membrane on the side of a small chamber in which the 
target hangs from a pivot. 



Figure 3.8. Two schematic RFB configurations showing the absorbing well (left) and suspended 
target (right) types. 



There is a 
radiation force 
of about 68 ug 
permW of 
acoustic power 

An angle 
correction 
may be 

necessary (e.g. 
for angled 
Doppler or 
scanning 
sector probes) 



Commercial 
RFBs are 
available. It is 
also possible 
to make an 
RFB 




29 



3 The acoustic output of diagnostic ultrasound scanners 



71 " a t surface" It i s possible to use an RFB to determine some forms of the TI. The TI "at surface" is 
can be measured calculated either from total acoustic power or from the power emanating from a 1cm 
square region of the transducer face. The use of an RFB with suitable mask allows this 
latter quantity to be measured. 



3.3.3 Measurement of temperature 

The safety standard IEC60601-2-37 (2007) limits the temperature of the transducer surface 
to less than 50 °C when running in air and to less than 43 °C when in contact with a 
phantom at 33 °C (for externally applied transducers) or at 37 °C (for internal transducers). 
It is often these temperature limits (rather than a limit on I or MI) that restrict the 
acoustic output of a transducer. 



T , in air can be 

surf 

measured most 
conveniently 
using an 
infra-red camera 



The most practical way to check compliance with the limit in air is to use an infared 
camera, which can be bought for between £1000 and £2000. Lower cost options are to use 
small wire thermocouples or single-point infra-red themometers but IR cameras have 
the advantage that the location of the hottest part of the transducer is visible (Figure 
3.9) so the measurement process becomes much faster and easier: the hottest part is not 
always in the middle of the transducer (Hekkenberg and Bezemer, 2004). In addition, 
thermocouples can perturb the temperature field leading to higher or lower values; and 
the spot-size on many infra-red thermometers is relatively large (1-2 mm), leading to 
spatial-averaging. The value of the surface emissivity is usually adjustable on the camera 
and should be set to give the correct temperature value when the transducer is "cold" i.e. 
at room temperature before ultrasound is applied. It is often instructive to see how the 
temperature distribution changes when the operating mode is changed or other scanner 
controls are adjusted. The measurement should be carried out over 30 min but it is often 
obvious long before that if the temperature is likely to approach the threshold value. 



f . on tissue is 

surf 

estimated using 
a phantom and 
a miniature 
thermocouple 



Measurement of surface temperature with the probe in contact with a phantom 
(Hekkenberg and Bezemer, 2004; Calvert and Duck, 2006) is more complicated but a 
phantom to mimic skin over soft tissue is available from the National Physical Laboratory. 
This consists of an agar gel covered with a layer of silicone rubber and meets the 
specification of IEC60601-2-37 (2007). A thermal sensor is not usually included, so users 
must supply their own: flat, metal film thermocouples have been widely used [e.g. type 
C02-K from Omega Engineering, Manchester, (UK)] . Thin wire thermocouples can also be 




Figure 3.9. Infra-red images of a linear array transducer operating in B-mode (left — maximum 
= 27.7°C), colour-flow (centre — maximum = 31.5°C) and PW Doppler mode (right — maximum 
= 31.6°C). 

30 



The acoustic output of diagnostic ultrasound scanners 3 



used but, in either case, it is preferable to avoid type T thermocouples since the thermal 
conductivity of copper is very high and distorts the temperature fi dd more than other 
types. Again, it is often instructive to observe how the temperature varies as the scanner 
controls are adjusted (Figure 3.10). 



Thermal phantoms can be made to mimic particular tissue paths and have an important 
role in evaluating any potential hazard arising from ultrasound-induced heating. 
Shaw et al. (2011) used a phantom designed to mimic the neonatal head to estimate the 
temperature rise at several locations in the head due to scanning through the fontanel 
at typical clinical settings. They found that approximately 35% of the configurations 
studied gave a temperature increase at the phantom skin surface in excess of 6 °C in less 
than 10 min. They also found that there was no useful correlation between the displayed TI 
and the temperature measured in the phantom: the average skin surface temperature on 
the phantom was 6 times larger than the average TI value. This is because the model for 
calculating TI completely ignores the self-heating of the transducer, which is actually the 
dominant factor governing T - The use of phantoms is not restricted to measurement 
of surface temperature (Shaw et al, 1998, 1999; IEC62306, 2006). 



7" surf on tissue is 
usually much 
higher than the 
TI value 



3.4 Control settings that give the highest output levels 

Awareness of the control settings that are likely to give the highest output levels is 
important both for users wishing to avoid high outputs and reduce the MI or TI value 
for safety reasons, and to measurers who are trying to maximize the output. Those who 
look for worst-case values must have an understanding of the operating principles of 
the particular machine, since the number of different possible combinations of control 
settings can run into millions. The nature and range of controls is constantly changing 
with the evolution of new scanning features, so provision of a rigid universal protocol is 
not possible. Controls on some of the newer machines can have quite unexpected effects, 
as manufacturers often arrange for, say, drive voltages or pulse repetition frequencies 
to change automatically when controls are set in a way which would otherwise cause a 
particular safety parameter, such as I or TI, to exceed a predetermined limit. However, 



Awareness of 
the control 
settings is 
important for 
users wishing 
to avoid high 
outputs and 
reduce the Ml 
or TI value for 
safety reasons 



B-mode — Move focus from Colour plus Set image depth 
harmonics off 10cm to 5cm PW Doppler to maximum 



TIS was 1 .9 here 




Figure 3.10. Example showing the variation in the surface temperature of a 3 MHz linear array 
transducer as scanner settings are adjusted. 

31 



3 The acoustic output of diagnostic ultrasound scanners 



Measurement 
of "worst- 
case" settings 
requires an 
understanding 
of the operating 
principles of 
the particular 
machine, since 
the number of 
different possible 
combinations of 
control settings 
can run into 
millions 

High power 
is often 
associated 
with longer 
focal depths 



published protocols can provide useful guidance (Henderson et ah, 1994; Whittingham, 
2000; IPEM, 2010) and help to reduce the search to manageable proportions. 

Apart from the obvious example of setting the output power control to maximum, 
there are a few general observations that can be made about producing high output 
levels. In general, the effects of controls on output levels depend upon the operating 
mode, but selecting a deep transmission focus often involves an increase in acoustic 
power whichever mode is chosen. This is because manufacturers often arrange for 
the transmission aperture to be increased if a deep transmission focus is selected, in 
order to maintain a narrow beam width and good sensitivity at depth. Apart from 
the effect of the larger aperture, the drive voltage applied to each element may also 
be increased to compensate for the greater attenuation anticipated for deep targets. 
Linear and curvilinear arrays give the greatest opportunity to vary the aperture. For 
smaller sector scanners and phased arrays, most of the aperture is used even for small 
depths. The power may not increase so much with display depth, because to do so 
would result in a higher energy density (and hence excessive temperature) at the 
transducer face. 



/ , can be 

spta 

increased 
when using 
the write- 
zoom control 
in B-mode, 
and with a 
narrow colour 
box in colour 
Doppler 
modes 



When operating in B-mode, the increase in power usually associated with a deep focus 
setting will also increase I . Activation of a write-zoom box is another way by which 
I is increased, particularly if the box is narrow. Unlike read-zoom, which simply 
magnifies part of the stored image, write-zoom involves a selected area being re- 
scanned at a higher line density. This leads to higher temporal-average intensities, since 
the probe continues to transmit the same energy per second, but this is restricted to a 
narrower area. Write-zoom may also lead to a higher pulse repetition frequency, since 
there is no need to wait for echoes from beyond the box, and this will increase temporal- 
average power and I even further. Since the transmission focus (or multiple foci) 
is usually automatically set to lie within the box, the highest I and power levels in 
B-mode are usually associated with a fairly deep and narrow zoom-box. A similar 
effect, only at generally higher output levels, also occurs with the colour box in colour 
Doppler modes, e.g. colour-flow mapping mode or colour Doppler power mode. Again, 
this may be less marked for smaller sector scanners and phased arrays than for linear 
or curvilinear arrays. It is sometimes possible to adjust the sector angle from a sector 
scanner or phased array. Reducing the angle can often increase the frame rate and so 
increase I and surface temperature. 



The position 
of the highest 
intensity in 
tissue may not 
be within the 
zoom or colour 
box itself 



Note that the position at which the I spta occurs is not generally within the zoom-box itself, 
but rather at a depth close to that of minimum slice thickness. Nevertheless, selection of a 
zoom-box still increases I , since it reduces the width of the scanned field at all depths, 
including that at the minimum slice thickness. Also note that, on some machines, if the 
write-zoom or colour box is moved to the very great depths, the power and J may not 
be as large as at a less extreme depth setting, since the aperture may not be able to expand 
any further, yet the pulse repetition frequency will be lower. 



In stationary beam modes, such as M-mode and spectral Doppler, temporal-average 
intensities are directly proportional to temporal-peak intensities, provided the pulse 

32 



The acoustic output of diagnostic ultrasound scanners 3 



repetition frequency (prf) remains constant. Thus, with this proviso, control settings 
that maximize p will be those that also maximize I . . A large p, and hence I . , is 

r r spta o f f spta 



usually produced if the operator-controlled focus (which acts in the scan plane) is set 
close to the (fixed) elevation focus, since this increases the strength of focussing in a 
three-dimensional sense. However, as discussed above, setting the focus (or range-gate 
in the case of spectral Doppler) to a greater depth may well increase the transmission 
aperture and drive voltage, and hence produce even greater pressures and intensities 
near the scan-plane focus. Only practical measurement can establish which of these two 
effects will produce the greatest p r and I 

In spectral Doppler mode, a high Doppler frequency scale setting, or the selection of 
"high prf" mode, is likely to produce a higher power and I . In themselves, these 
prf-related controls would not be expected to affect p r values, but manufacturers 
sometimes arrange for drive voltages, and hence pressure amplitudes, to be reduced 
for safety reasons if a high prf is selected. A short range-gate is likely to give higher 
p r values, since drive voltages are usually reduced as gate length increases, again for 
safety reasons. The effect of range-gate length on power and J s ta is difficult to predict, 
for the same reason. 

The use of harmonic imaging modes is often accompanied by higher p (to generate more 
non-linearity) leading to greater output power and higher 7 sp[a . 



3.5 Acoustic output values 

3.5.1 Independent measurements of acoustic outputs 

Measurements of acoustic exposure parameters from diagnostic ultrasound systems 
have long been of interest for assessing their acoustic safety. The first surveys of acoustic 
outputs to include real-time B-scan array systems (c.f . static B-scanners) were published 
in 1978 (Carson et al, 1978) and 1985 (Duck et al, 1985), when this technology was 
relatively new. Although the number of array systems studied in these reports is small 
(two and four respectively) and some of the measurement methods differ from those 
now used, the values are of interest because they are so much lower than those reported 
more recently for current systems. These early reports are discussed in more detail later. 

The most comprehensive surveys of acoustic output values for ultrasound systems using 
real-time transducers were published or carried out in the 1990s (Duck and Martin, 1991; 
Henderson et al., 1995; Whittingham, 2000). These surveys were carried out in the UK 
by NHS medical physics departments, independently of the equipment manufacturers 
and relate to systems in active clinical use at the time of measurement. The measurement 
methods used were based on the principles described above, i.e. using a PVDF membrane 
hydrophone in a water bath to measure acoustic pressure and a RFB to measure acoustic 
power. In all 3 surveys, the active element of the hydrophone used was 0.5 mm in diameter. 

In these surveys, the spatial-peak values of pressure and the intensity parameters given 
were those measured in water at the point in the acoustic field where they achieved their 

33 



The highest 
p r in spectral 
Doppler mode 
is commonly 
associated with 
the shortest 
range-gate, 
and lowest prf. 
A high Doppler 
frequency scale 
setting, or 
the selection 
of "high prf" 
mode, is likely 
to produce a 
higher power 
and L 



Several surveys 
of acoustic 
exposure 
parameters 
were carried 
out in the 
1990s. These 
reported 
peak values 
of pressure 
and intensity 
measured in 
water under 
worst-case 
conditions 



3 The acoustic output of diagnostic ultrasound scanners 



Mean and 
median values 
of p r from the 
three 1 990s 
surveys in 
3 modes of 
operation are 
of the order 
of 2.4MPa, 
with much 
overlap in 
the ranges of 
values, but a 
slow upward 
trend 



maximum value. No derating of values, as used in the measurement and calculation of 
safety indices (see above) was applied. In each case, the ultrasound system controls, for 
the particular mode of operation (e.g. B-mode or colour Doppler), were manipulated to 
achieve the highest value of the parameter of interest (e.g. p r or I J, i.e. the worst-case 
value. As described above, the operating conditions and hydrophone locations required 
to give worst-case values of p r and I in each mode of operation are likely to be quite 
different. 

For all three surveys, measurement data from all types of real-time transducer 
(linear/curved array, phased array, mechanical, transcutaneous and intracavity) are 
combined. The survey by Duck and Martin (1991) hcluded data from 108 real-time 
transducers and 44 scanners from 19 manufacturers. Pulsed Doppler measurements were 
from 17 systems from 1 1 manufacturers. Henderson et al. (1995) studied 82 scanners and 223 
transducers from 18 manufacturers. The data from Whittingham (2000) related to similar 
measurements made in the period 1995-1998. The survey by Duck and Martin (1991) gave 
values for peak compression and rarefaction pressures, 7 spta , I g a and acoustic power. 
As the 2 later surveys gave values only for peak rarefaction pressure, I s u and acoustic 
power, values for these 3 parameters are reviewed here to allow identification of trends. 

Table 3.1 gives the range, median and mean values of peak rarefaction pressure quoted by 
the three surveys above in real-time B-mode, pulsed Doppler and colour Doppler modes. 



Values of 
/ . from the 

spta 

1 990s surveys 
were highest 
in pulsed 
Doppler mode 
(mean>1 W 
cnr 2 ) and least 
in B-mode. 
There was a 
strong upward 
trend in / , in 

spta 

B-mode but 
little change in 
pulsed Doppler 
values 



Values for peak rarefaction pressure for all modes and surveys are typically in the range 
0.5-5 MPa with mean and median values of the order of 2.4 MPa. For each individual 
survey, there is much overlap in the ranges of pressure values for the three modes of 



Table 3.1. Worst-case values of peak rarefaction pressure (MPa) as measured in water. Data 
taken from Duck and Martin (1991), Henderson era/. (1995) and Whittingham (2000) (1998 
survey). The number of transducers for which measurements are included is n. 





1991 


1995 


1998 


B-mode 








Range 


0.6^.3 


0.4-5.5 


0.5^.6 


Median 


2.1 


2.4 


2.4 


Mean 


2.1 


2.4 


2.6 


n 


108 


190 


100 


Pulsed Doppler 








Range 


0.2-3.8 


0.7-5.3 


0.6-5.5 


Median 


1.6 


2.1 


2.4 


Mean 


1.8 


2.2 


2.4 


n 


42 


118 


82 


Colour Doppler 








Range 


0.9-3.9 


0.5^.2 


0.8^.9 


Median 


2.3 


2.4 


2.6 


Mean 


2.4 


2.4 


2.8 


n 


18 


87 


79 



34 



The acoustic output of diagnostic ultrasound scanners 3 



operation, although in the 1991 survey, the mean and median values in pulsed Doppler 
mode are somewhat lower. Comparison of values across the three surveys shows a 
gradual trend to higher values and less difference between modes of operation. 

Table 3.2 gives values for 7 spta from the 1991, 1995 and 1998 surveys in B-mode, pulsed 
Doppler and colour Doppler. 

Values for 7 spta in all modes and surveys show much wider ranges of values than for 
peak rarefaction pressure. There are also substantial differences in mean and median 
values for the 3 modes of operation. In the 1991 survey, there is a clear progression 
in mean values from B-mode to colour Doppler mode to pulsed Doppler mode, with 
almost an order of magnitude increase between modes. Mean and median values of 
7 spta in pulsed Doppler are greater than 1 Wcnr 2 . High values of I s ta would be expected 
in pulsed Doppler mode due to the use of a stationary beam, whereas in B-mode, the 
beam is scanned across the full image and in colour Doppler mode across the region 
of the colour box. Across the surveys, there is a strong upward trend in 7 spta values in 
B-mode and to a lesser extent in colour Doppler mode, but no clear trend in pulsed 
Doppler mode. 

Table 3.3 gives values for acoustic power from the 1991, 1995, and 1998 surveys in 
B-mode, pulsed Doppler and colour Doppler. Values for acoustic power show relatively 
small differences between modes with typical mean values being of the order of 100 mW. 
Typical values for the 1995 and 1998 surveys in B-mode and pulsed Doppler mode are 
higher than those from the 1991 survey, especially in B-mode. 



Values of 
acoustic power 
from the 1990s 
surveys showed 
small differences 
between modes 
with mean 
values in colour 
Doppler and 
pulsed Doppler 
of the order of 
100mW 



Table 3.2. Worst-case values of spatial-peak temporal-average intensity (mWcrrr 2 ) as mea- 
sured in water. Data taken from Duck and Martin (1991), Henderson ef a/. (1995) and 
Whittingham (2000) (1 998 survey). The number of transducers for which measurements are 
included is n. 





1991 


1995 


1998 


B-mode 








Range 


0.3-177 


0.3-991 


4.2-600 


Median 


6.0 


34 


94 


Mean 


17 


106 


175 


n 


101 


194 


100 


Pulsed Doppler 








Range 


110^520 


173-9080 


214-7500 


Median 


1140 


1180 


1420 


Mean 


1380 


1659 


1610 


n 


42 


118 


82 


Colour Doppler 








Range 


25-511 


21-2050 


27-2030 


Median 


96 


290 


330 


Mean 


148 


344 


470 


n 


19 


87 


79 



35 



3 The acoustic output of diagnostic ultrasound scanners 



Table 3.3. Worst-case values of acoustic power (mW). Data taken from Duck and Martin 
(1991), Henderson et al. (1995) and Whittingham (2000) (1998 survey). The number of 
transducers for which measurements are included is n. No values for acoustic power in 
colour Doppler mode were given in the 1991 survey. 





1991 


1995 


1998 


B-mode 










n S3-350 


0.3-285 


4-256 




7.1 


7S 




Mean 


19.1 


77.8 


64 


n 


51 


45 


29 


Pulsed Doppler 








Range 


8.7-210 


10-440 


11-324 


Median 


42 


100 


129 


Mean 


63.5 


124 


144 


n 


20 


39 


22 


Colour Doppler 








Range 




15^40 


35-295 


Median 




90 


118 


Mean 




119 


138 


n 




29 


22 



3.5.2 Manufacturer declared acoustic outputs 

Although, to the knowledge of the authors, no further independent surveys of acoustic 
outputs from clinical ultrasound systems have been published since 2000, acoustic 
output data for more recent models are available from equipment manufacturers. As 
part of the regulatory systems for medical ultrasound devices in Europe and the USA, 
manufacturers must declare acoustic output values to demonstrate the acoustic safety 
of their equipment (see Chapter 10). In the USA, the FDA requires manufacturers to 
declare maximum values of MI and TI for each transducer and mode of operation and 
demonstrate that these are within prescribed limits (FDA, 2008). The format of the 
declaration is essentially the same as that in the international standard IEC60601-2-37 
(2007). For medical ultrasound systems to be placed on the market in the European 
Union, manufacturers must demonstrate compliance with the essential safety req- 
uirements of the European Medical Devices Directive (European Communities, 1993). 
Acoustic safety may be demonstrated by declaring acoustic output values in compliance 
with IEC60601-2-37 (2007) or IEC61157 (2007) although no upper limits are enforced. 
IEC61157 includes a requirement to declare maximum values for peak rarefaction 
pressure and 7 spta for each transducer and mode of operation. These must be measured 
in water at the location where they achieve maximum value, with no derating applied 
and hence may be compared with the independent survey values discussed in the 
previous section. Martin (2010) has reported values of peak rarefaction pressure (p r ) 
and I , from such IEC61157 declarations for ultrasound systems on the market in 2008. 

spta J 

These are shown in Table 3.4. 

36 



The acoustic output of diagnostic ultrasound scanners 3 



Table 3.4. Manufacturer declared values for maximum in-water peak rarefaction pressure (p r ) 
and / from Martin (2010). The number of transducers for which measurements are included 
is n. 





fJ r V lvlrd / 


/ (mW rm -2 ) 
'spta U ' IVV Cm 1 


B-mode 






Range 


2.3-6.4 


20-1100 


Median 


3.7 


273 


Mean 


3.9 


341 


n 


79 


79 


Pulsed Doppler 






Range 


2.1-6.7 


271-2830 


Median 


4.2 


749 


Mean 


4.2 


860 


n 


79 


79 


Colour Doppler 






Range 


1.4-6.7 


51-1480 


Median 


4.2 


450 


Mean 


4.1 


466 


n 


79 


79 



In this manufacturer declared data, mean and median values of p are of the order of 
4MPa, with much overlap between modes in the ranges of values. The highest mean and 
median values are in colour Doppler and pulsed Doppler modes rather than B-mode. 
There is still a progression in mean and median values for I from B-mode to colour 
Doppler to pulsed Doppler but the differences between the modes are much smaller than 
those in the surveys reported earlier. In the declared data, the mean J spta value in pulsed 
Doppler is approximately 2.5 times that in B-mode, whereas in the 3 independent surveys, 
this ratio varies between 9.2 and 81. 

As described earlier, it is now more common to characterize the acoustic output of 
ultrasound systems in terms of safety indices, which indicate the risk of cavitation or 
heating, rather than using measured exposure parameters such as pressure and intensity. 
Martin (2010) reported declared values of MI and TI from four ultrasound equipment 
manufacturers. These are maximum values achievable under worst-case operating 
conditions. 



Mean values 
of p r declared 
recently by 
manufacturers 
are of the order 
of 4 M Pa, with 
much overlap 
between 
modes. 

Declared values 
of / , were 

spta 

highest in 
pulsed Doppler 
but with much 
more overlap 
between 
modes than in 
earlier surveys 



Under FDA rules, the maximum permitted value for MI is 1.9 and the maximum 
measured value plus the uncertainty in the measurement must be within this limit. For 
TI, the normal maximum value is 6.0. 

Table 3.5 gives manufacturer declared values of MI (combined from 4 manufacturers), 
TIS, TIB and TIC in B-mode and pulsed Doppler from Martin (2010). The ranges of 
MI values for B-mode and pulsed Doppler are the same, consistent with the large 
overlap in ranges for the declared values of p r in Table 3.4. Maximum MI values are 
less than the permitted maximum of 1.9 due to the need to allow for measurement 

37 



Maximum values 
of Ml declared 
by manufacturers 
are of the order 
of 1 .25 (mean) 
in B-mode and 
pulsed Doppler 
with ranges 
extending to 1 .7 



3 The acoustic output of diagnostic ultrasound scanners 



Table 3.5. Declared values of Ml (combined manufacturer values), TIS, TIB and TIC in B-mode 



and pulsed Doppler mode from Martin (2010). 




Ml 


TIS 


TIB 


TIC 


B-mode 










Range 


0.2-1.7 


0.5-4.1 


0.02-4.0 


0.1-5.9 


Mean 


1.25 


0.99 


0.78 


1.7 


n 


177 


131 


86 


92 


Pulsed Doppler 










Range 


0.2-1.7 


0.08-5.0 


0.26-7.0 


0.15-6.6 


Mean 


1.23 


1.1 


2.3 


1.8 


n 


147 


138 


145 


87 



Mean 
manufacturer 
declared 
values of Tl are 
generally much 
lower than the 
FDA limit of 6.0 



uncertainty. The relatively high mean values of MI (of the order of 1.25) suggest that 
many transducers are capable of producing an MI near to the FDA regulatory limit. 

Mean declared values of TIS, TIB and TIC in both modes are much lower than the 
normal maximum FDA value of 6.0, showing that many transducers are not capable 
of producing this value. For some transducers, this may be due to the transducer 
temperature reaching the regulatory limit before a high TI value is achieved. Mean 
declared values of TIS and TIB in B-mode and TIS in pulsed Doppler mode are all of the 
order of 1.0, reflecting the fact that the same mathematical model is used for TIS and 
TIB for scanned modes (e.g. B-mode) and TIS in unscanned modes for apertures <1 cm 2 
(e.g. pulsed Doppler). Mean values for TIC in both modes and TIB in pulsed Doppler 
mode are higher. 



Maximum values 
of Ml reported 
during clinical 
use cover a 
similar range to 
manufacturer 
declared 
maximum values 
but are lower on 
average 



ter Haar (2008) reported values for TI and MI obtained from a survey of ultrasound users 
in the UK. These values were reported in 2 categories: (i) the default values obtained 
at switch-on or with each new patient and (ii) the maximum values achieved during 
an individual patient scan. Hence the values from this survey represent typical MI 
and TI values found in clinical use rather than worst-case system values. In this study, 
values were reported according to the clinical application of the scan rather than the 
mode of operation. Of the 48 abdominal scans reported, 44 were B-mode or B-mode with 
harmonic imaging and 4 included use of colour Doppler. TI values were reported as TIS 
for 36 scans and TIB for 5. In the obstetric category, 66 scans were reported as B-mode, 30 
as colour Doppler and 19 included pulsed Doppler (other reports did not specify mode 
of operation). TI values were reported as TIS for 20 scans, TIB for 72 scans and TIB/TIS/ 
TIC for 31. 

Table 3.6 gives the maximum values of MI and TI recorded during abdominal and 
obstetric scans. Although the equipment used in this study was not the same as that for 
manufacturer declared values, both studies related to equipment that was available in 
2008. The large overlap in the ranges of the manufacturer and clinical values of MI suggest 
that the full range of available MI values may be used in clinical practice. Mean values 
of maximum MI used in clinical practice are of the order of 60-75% of the maximum 

38 



The acoustic output of diagnostic ultrasound scanners 3 



Table 3.6. Maximum values of Ml and Tl recorded during abdominal and obstetric scans by 
clinical users (from ter Haar, 2008). 





Ml 

max 


Tl 

max 


Abdominal 






Range 


0.4-1.6 


0.1-0.8 


Mean 


0.97 


0.56 


n 


48 


46 


Obstetric 






Range 


0.2-1.6 


0.1-2.5 


Mean 


0.74 


0.98 


n 


220 


167 



available. Direct comparison between clinical and manufacturer declared Tl values is 
more difficult due to the mixture of modes of operation and types of Tl reported in the 
clinical study. However, the range and mean Tl values reported for abdominal scans 
were significantly lower than the worst-case manufacturer Tl values. Maximum Tl values 
reported for obstetric applications were higher than for abdominal scans, presumably 
due to the use of pulsed Doppler in many of the cases. 



Maximum Tl 
values reported 
for abdominal 
scans are 
lower than the 
maximum B-mode 
values declared by 
manufacturers 



3.5.3 Trends in acoustic outputs 



The acoustic output surveys from the 1990s described above showed gradual increases 
through that decade in mean and median values of peak rarefaction pressure in B-mode, 
colour Doppler and pulsed Doppler. Mean and median values of I a ta in B-mode 
and colour Doppler increased more strongly. More recent trends can be revealed by 
comparing values from these surveys to manufacturer declared values. In making this 
comparison, it must be taken into account that the measurement equipment used and 
measurement uncertainties may not be the same. Also the manufacturer values may 
represent the maximum from a small batch of the same type of transducer and system 
model. Figure 3.11 shows graphically the range and mean values for peak rarefaction 
pressure from the 1990s surveys and the manufacturer declared values (from Martin, 
2010). Mean manufacturer declared values of p r were of the order of 4MPa, representing 
an increase of the order of 50% in mean B-mode and colour Doppler values and 75% in 
pulsed Doppler since 1998. 

Figure 3.12 shows a similar comparison of survey and manufacturer declared values of 
J ta in the three modes of operation. In B-mode, the upward trend observed in the 1990s 
surveys has continued. The mean value in 2010 (341mWcirr 2 ) was almost twice that 
measured in 1998 (175mWcmr 2 ). However, in pulsed Doppler mode, a downward trend 
is seen; the mean declared value of 7 spta was approximately half that reported in 1998. In 
colour Doppler mode, the mean 7 spta value was similar to that reported in 1998. A further 
important observation which emerges from this comparison is that the opposing trends 
in B-mode and pulsed Doppler has resulted in much overlap in the ranges of I values 
for the three modes of operation. It is no longer the case that I values in pulsed Doppler 
are many times greater than in B-mode. 

39 



Peak rarefaction 
pressures declared 
by manufacturers 
(2010) are 50-75% 
higher than those 
reported in a 1998 
measurement 
survey 



Maximum / , 

spta 

values declared 
by manufacturers 
(2010) showed 
an increase for 
B-mode but a 
reduction for 
pulsed Doppler 
mode compared 
to 1 998 survey 
values, resulting 
in increased 
overlap in values 
between modes 



3 The acoustic output of diagnostic ultrasound scanners 




1991 



1995 



1998 



2010 



Year of Survey 



Figure 3.1 1. Manufacturer declared values of peak rarefaction pressure (p r ) in (2010) compared 
to 3 previous surveys in B-mode, pulsed Doppler and colour Doppler (from Martin, 2010). 



3500 
3000 
2500 
o 2000 

E, 

J 1500 
1000 
500 
0 



| B-mode | Pulsed Doppler |J Colour Doppler 



A4520 



1991 



A9080 
i 



_ 



1995 



A7500 
i 



1998 



U 1 



2010 



Figure 3.12. Manufacturer declared values of spatial-peak temporal-average intensity (/ ) in 
(2010) compared to 3 previous surveys in B-mode, pulsed Doppler and colour Doppler mode 
(from Martin, 2010). 



40 



The acoustic output of diagnostic ultrasound scanners 3 



3.6 Discussion 



In previous years, those carrying out output measurements on scanners have tended 
to concentrate on finding the settings which give the maximum output. This is still 
appropriate for regulatory purposes but, with the extreme complexity of modern 
scanners, it is debatable whether it is worth the effort of trying to identify these settings 
in a hospital environment. They may not even be particularly relevant to hospital users 
and patients. There are two alternative approaches which seem to be more efficient and 
relevant outside of the regulatory environment. The first is to measure the presets. Almost 
all ultrasound exams start with the selection of a clinical "preset" appropriate for the 
type of scan being undertaken and, although practice varies between individuals, it is 
common that relatively few settings that affect the output (such as display depth and scan 
mode) are adjusted during the scan. So the output of the preset is a fair representation 
of the output during a typical scan of that type. The use of the preset also provides a way 
of reproducing the output configuration for instance to look for changes over time or 
following hardware or software upgrades: some presets could be "locked" and dedicated 
to testing. Testing could involve hydrophone measurements but even simpler testing 
with an RFB or a thermal phantom would provide useful tests of consistency. 



It may be 
more useful to 
measure the 
clinical presets 
and to validate 
displayed 
indices than to 
try to maximize 
output 



A second approach is to validate the displayed MI and TI values. Limited validation 
can be carried out using an RFB to measure power, although more thorough validation 
requires hydrophone measurements. Determination of MI and TI requires a measurement 
of power and measurements of p r and I along the beam-axis. There is a strong case for 
saying that every medical physics department with an ultrasound responsibility should 
be capable of validating the displayed indices and should do this as part of acceptance 
testing when scanners are purchased. Users are expected to make risk-benefit judgements 
based on the values displayed and need to be reassured that they are reliable. A system 
targeted to making only these measurements could be more compact and economical than 
a more general hydrophone system. A prototype of such a system has been assembled at 
the National Physical Laboratory. 



To ensure the safe use of ultrasound in diagnosis, it is important for clinical users to have 
some awareness and understanding of the acoustic output of their equipment and how 
it is affected by the way in which they operate it. Review of reported values for acoustic 
output parameters shows that maximum values for some parameters may now be more 
than 2 orders of magnitude greater than those reported for the first real-time B-mode 
transducers. Carson et al. (1978) reported a maximum value for I of 15.9 mW cm -2 and 
Duck et al. (1985) a maximum value of 2.5mWcm~ 2 for early linear array transducers 
operating in B-mode. In 2010, a worst-case value in real-time B-mode of HOOmWcm -2 
was declared for this parameter by an ultrasound system manufacturer (Martin, 2010). 
While the developments in ultrasound technology that have taken place over this time 
scale have resulted in huge improvements in image quality and system performance, it 
is now clear that they have also resulted in some significant increases in acoustic output 
levels, particularly in B-mode. 



Maximum 
values for some 
parameters are 
more than 100 
times greater 
than those 
reported for the 
first real-time 
B-mode scanners 



41 



3 The acoustic output of diagnostic ultrasound scanners 



Some new 
imaging 
methods 
require higher 
output: 
regulations 
may eventually 
be modified 
to allow more 
effective use 
of emerging 
techniques 



The review given above of acoustic output surveys from the 1990s and more recent 
manufacturer declared outputs showed steady increases in the mean values of peak 
rarefaction pressure in B-mode, pulsed Doppler and colour Doppler. Such increases would 
be expected from manufacturers' efforts to improve transducer efficiency and focussing 
methods, thus allowing the use of higher frequencies for a given application. Acoustic 
pressure levels are now effectively regulated by the FDA through the maximum permitted 
values of MI. MI increases with p r but reduces with the square root of frequency and there 
is evidence (Martin, 2010) that some manufacturers are working to this regulatory limit, 
using higher p values at higher frequencies. 

More recent developments in technology may have influenced acoustic pressure and 
MI values. For example, harmonic imaging requires the use of high acoustic pressures 
to generate harmonics in the pulse waveform via non-linear propagation. Harmonic 
imaging (Tranquart et ah, 1999) may also be associated with higher MI values, as some 
harmonic imaging methods also require the transmit frequency to be reduced. Shear wave 
techniques (Bercoff et ah, 2004) must make use of the highest permissible output levels to 
generate acoustic radiation stress in tissues. It is possible that existing regulations may be 
modified to permit the use of higher outputs. 

The largest changes in acoustic output values over the last 20 years have been in I 
levels for real-time B-mode. In the 1991 survey, the mean value of this parameter in 
B-mode was smaller than that in pulsed Doppler by a factor of 84. This would be 
expected due to the use of a stationary beam in pulsed Doppler mode and the use of 
relatively long pulses and high prf. However, in the manufacturers' declared values 
from 2010, this ratio was reduced to 2.5. While there had been a reduction in the mean 
value of 7 spta in pulsed Doppler mode, the mean value in B-mode had increased from 
17 to 341mWcm~ 2 . The 2010 data also showed much overlap in the ranges of values 
for these 2 modes. 

Values for I in B-mode have presumably increased over this time due to 
improvements in transducer efficiency and focussing methods. New pulse sequencing 
techniques designed to increase frame rates and the use of longer coded pulses (Chiao 
and Hao, 2005) would also contribute to this increase. Corresponding increases in 
the already high values in pulsed Doppler would be restricted by FDA limits on I 
(Chapter 10). 



BMUS 
recommends 
that obstetric 
scanning time 
is limited for Tl 
more than 0.7 



Maximum achievable TI values declared by manufacturers, on average, are low compared 
to the FDA normal maximum permitted value of 6.0. Mean values of TIS and TIB in 
B-mode and TIS in pulsed Doppler mode are of the order of 1.0 or less. The mean TIB 
value declared for pulsed Doppler (2.3) was more than twice this value. Mean values 
for TIC in B-mode and pulsed Doppler were 1.7 and 1.8. While these values may seem 
low in relation to the maximum FDA value, they are still significant in acoustic safety 
terms. For obstetric and neonatal transcranial scanning, the BMUS safety guidelines 
(BMUS, 2010) recommend that scanning time is restricted for any TI value greater than 
0.7. For a TIB of 2.3, it is recommended that the duration of such scans be restricted to 
4min. Maximum TI values used in obstetric scanning as reported by ultrasound users 

42 



The acoustic output of diagnostic ultrasound scanners 3 



(ter Haar, 2008) are more reassuring. The majority of TI values used in obstetric scanning 
were less than 0.7 with a mean value of 0.44. Of the 231 obstetric scans reported, only 
three were identified that did not conform to the BMUS guidelines in terms of TI level 
and scanning time. 

Acknowledgements 

The authors would like to acknowledge the contribution of Dr Tony Whittingham who 
wrote the equivalent chapter in the second edition of this book. We have reused some 
parts of his original chapter. 

References 

Bercoff J, Tanter M, Fink M. 2004. Supersonic shear imaging: a new technique for soft 
tissue elasticity mapping. IEEE Trans Ultrason Ferroelectr Freq Control, 51, 396-409. 
BMUS. 2010. Guidelines for the Safe Use of Diagnostic Ultrasound Equipment. Available 
at: bmus.org. 

Calvert J, Duck FA. 2006. Self-heating of diagnostic ultrasound transducers in air and in 
contact with tissue mimics. Ultrasound, 14, 100-108. 

Carson PL, Fischella PS, Oughton TV. 1978. Ultrasonic powers and intensities produced 

by diagnostic ultrasound equipment. Ultrasound Med Biol, 3, 341-350. 

Chiao RY, Hao X. 2005. Coded excitation for diagnostic ultrasound: a system developers 

perspective. IEEE Trans Ultrason Ferroelectr Freq Control, 52, 160-170. 

Duck FA, Martin K. 1991. Trends in diagnostic ultrasound exposure. Phys Med Biol, 36, 

1423-1432. 

Duck FA, Starritt HC, Aindow JD, Perkins MA, Hawkins AJ. 1985. The output of pulse- 
echo ultrasound equipment: a survey of powers, pressures and intensities. Br J Radiol, 58, 
989-1001. 

European Communities. 1993. Council directive 93/42/EEC of June 1993 concerning 
medical devices. Off J Eur Communities, 169. 

Farmery MJ, Whittingham TA. 1978. A portable radiation-force balance for use with 
diagnostic ultrasonic equipment. Ultrasound Med Biol, 3, 373-379. 

FDA. 2008. Information for Manufacturers Seeking Marketing Clearance of Diagnostic 
Ultrasound Systems and Transducers. Rockville, MD: Food and Drug Administration. 
Hekkenberg RT, Bezemer RA. 2004. On the development of a method to measure the 
surface temperature of ultrasonic diagnostic transducers. / Phys Conf Ser, 1, 84. doi: 
10.1088/1742-6596/1/1/019. 

Henderson J, Jago J, Willson K, Whittingham TA. 1994. Development of protocols for 
measurement of maximum spatial peak temporal average intensity from scanners 
operating in pulsed Doppler and colour Doppler modes. Br } Radiol, 67, 716. 
Henderson J, Willson K, Jago J, Whittingham TA. 1995. A survey of the acoustic outputs 
of diagnostic ultrasound equipment in current clinical use in the Northern Region. 
Ultrasound Med Biol, 21, 699-705. 

Hodnett M, Zeqiri B. 2009. A novel sensor for determining ultrasonic intensity. In 
Proceedings of the 38th Annual Symposium of the Ultrasonic Industry Association (UIA), 
Muratore R, Hodnett M (editors), doi: 10.1109/UIA.2009.5404029. 

43 



3 The acoustic output of diagnostic ultrasound scanners 

IEC60601-2-37 Edition 2. 2007. Medical Electrical Equipment - Part 2-37: Particular 
Requirements for the Basic Safety and Essential Performance of Ultrasonic Medical 
Diagnostic and Monitoring Equipment. Geneva, Switzerland: International Electrotechnical 
Commission. 

IEC61157 Edition 2.0. 2007. Standard Means for the Reporting of the Acoustic Output 
of Medical Diagnostic Ultrasonic Equipment. Geneva, Switzerland: International 
Electrotechnical Commission. 

IEC61161 Edition 2.0. 2006. Ultrasonics - Power Measurement - Radiation Force 
Balances and Performance Requirements. Geneva, Switzerland: International Electro- 
technical Commission. 

IEC62127-1 Edition 1.0. 2007. Ultrasonics - Hydrophones - Part 1: Measurement and 
Characterization of Medical Ultrasonic Fields up to 40 MHz. Geneva, Switzerland: 
International Electrotechnical Commission. 

IEC/TS62306 Edition 1.0. 2006. Ultrasonics - Field Characterisation - Test Objects for 
Determining Temperature Elevation in Diagnostic Ultrasound Fields. Geneva, Switzerland: 
International Electrotechnical Commission. 

IEC62359 Edition 2.0. 2010. Ultrasonics - Field Characterization - Test Methods for 
the Determination of Thermal and Mechanical Indices Related to Medical Diagnostic 
Ultrasonic Fields. Geneva, Switzerland: International Electrotechnical Commission. 
IPEM Report 102. 2010. Quality Assurance of Ultrasound Imaging Systems, Stephen 
Russell (editor). York, UK: Institute of Physics and Engineering in Medicine. 
Jago J, Henderson J, Whittingham TA, Willson K. 1995. How reliable are manufacturers' 
reported acoustic output data? Ultrasound Med Biol, 21, 135-136. 

Lewin PA, Ziskin MC (editors). 1992. Ultrasonic Exposimetry. Boca Raton, FL: CRC Press. 
Martin K. 1986. Portable equipment and techniques for acoustic power output and intensity 
measurement. In Physics in Medical Ultrasound, Evans JA (editor). London, UK: IPEM. 
Martin K. 2010. The acoustic safety of new ultrasound technologies. Ultrasound, 18, 
110-118. 

Perkins MA. 1989. A versatile force balance for ultrasound power measurement. Phys Med 
Biol, 34, 1645-1651. 

Preston RC. 1988. The NPL ultrasound beam calibrator. IEEE Trans Ultrasonics Ferroelectr 
Freq Control, 35, 122-138. 

Preston RC (editor). 1991. Output Measurements for Medical Ultrasound. London, UK: 
Springer- Verlag. 

Shaw A, Memoli G, Duck FA, Osborne J, Robinson H, Bartle D, et al. 2011. Neonatal 
Transcranial Ultrasound - Thermal Hazard and Risk Assessment. NPL Report. 
Teddington, UK: National Physical Laboratory. 

Shaw A, Pay NM, Preston RC. 1998. Assessment of the Likely Thermal Index Values for 
Pulsed Doppler Ultrasonic Equipment - Stage II: Experimental Assessment of Scanner/ 
Transducer Combinations. NPL Report CMAM 12. Teddington, UK: National Physical 
Laboratory. 

Shaw A, Pay NM, Preston RC, Bond AD. 1999. A proposed standard thermal test object 
for medical ultrasound. Ultrasound Med Biol, 25, 121-132. 

Shaw A, Preston RC. 1995. The NPL ultrasound beam calibrator: cost effective 
measurements to IEC61157. In Proceedings of the World Congress on Ultrasonics, 
Herbertz J (editor), pp. 923-926. 

44 



The acoustic output of diagnostic ultrasound scanners 3 

Sutton Y, Shaw A, Zeqiri B. 2003. Measurement of ultrasonic power using an acoustically 
absorbing well. Ultrasound Med Biol, 29, 1507-1513. 

Szabo TL. 2004. Ultrasonic exposimetry and acoustic measurements. In Diagnostic 
Ultrasound Imaging: Inside Out, Szabo TL (editor). Burlington, MA: Elsevier Academic 
Press. 

ter Haar G. 2008. Results of a survey of exposure conditions used in ultrasound scans in 
the U.K., February 2007. Ultrasound, 16, 110-113. 

ter Haar G, Shaw A, Pye S, Ward B, Bottomley F, Nolan R, et al. 201 1. Guidance on reporting 
ultrasound exposure conditions for bio-effects studies. Ultrasound Med Biol, 37, 177-183. 
Tranquart F, Grenier N, Eder V, Pourcelot L. 1999. Clinical use of ultrasound tissue 
harmonic imaging. Ultrasound Med Biol, 25, 889-894. 

Whittingham TA. 2000. The acoustic output of diagnostic machines. In The Safe Use of 
Ultrasound in Medical Diagnosis, ter Haar G, Duck FA (editors). 2nd Edition. London, 
UK: BMUS/BIR. 

Wilkens V. 2010a. A thermal technique for local ultrasound intensity measurement: part 1. 
Sensor concept and prototype calibration. Meas Sci Technol, 21, 115805 (8 pp.). 
Wilkens V. 2010b. A thermal technique for local ultrasound intensity measurement: 
part 2. Application to exposimetry on a medical diagnostic device. Meas Sci Technol, 21, 
115806 (10 pp.). 

Zeqiri B, Bickley CJ. 2000. A new anechoic material for medical ultrasonic applications. 

Ultrasound Med Biol, 26, 481-485. 

Zeqiri B, Zauhar G, Hodnett M, Barrie J. 2011. Progress in developing a thermal method 
for measuring the output power of medical transducers that exploits the pyroelectric 
effect. Ultrasonics, 51, 420^24. 



45 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 4 

Ultrasound-induced heating and 
its biological consequences 

Charles C. Church 1 and Stanley B. Barnett 2 

1 National Center for Physical Acoustics, University of Mississippi, Oxford, MS, USA 
2 Mona Vale, NSW, Australia 



Summary 

• Developing tissues of the embryo and foetus are particularly susceptible to damage 
by heating, and the effects can have serious consequences. 

• The induction of teratogenic effects depends on a combination of the elevation 
above normal physiological temperature and the duration for which the increased 
temperature is maintained. 

• A threshold dose of 0.5 min exposure to an increase of 4°C above normal body 
temperature may be hazardous to embryonic and foetal development. 

• In late pregnancy, the heated volume is small compared with the size of the foetus, 
and consequent biological effects may be difficult to detect unless a major neural 
pathway is perturbed. 

• Simple grey scale B-mode imaging is not capable of producing harmful temperature 
increases in tissue. 

• A diagnostic exposure that produces a maximum temperature rise of 1.5 °C above 
normal physiological levels (37 °C) does not appear to present a risk from thermal 
effects in humans for an imaging session of less than 30 min. 

• There are uncertainties in predicting in situ temperature increases in the embryo 
and foetus, and it is therefore prudent to use the minimum output consistent with 
obtaining the required diagnostic information. 

4.1 Introduction 

The normal human core temperature is generally accepted to be 37 °C with a diurnal 
variation of ± 0.5-1 °C (Mellette et al, 1951; Hardy, 1961), although 36.8 ± 0.4 °C may be 
the true mean for large populations (Mackowiak et al., 1992). Temperature in the human 
foetus is higher than maternal core body temperature by 0.3-0.5 °C during the entire 
gestation (Asakura, 2004), but in the third trimester (near-term) the temperature of the 

46 



Ultrasound-induced heating and its biological consequences 4 



foetus is higher by 0.5 °C than that of its mother (Macaulay et al, 1992). Biological tissues 
absorb ultrasound, resulting in heating above normal physiological temperatures, and an 
increase in temperature of sufficient magnitude and duration can damage or kill biological 
tissues. These facts are well known to most professional users of diagnostic ultrasound 
equipment. However, the details of the relationship between exposure to heat and the 
resulting effects are less well understood, even by those who use ultrasound every day. 



Temperature fundamentally affects biochemical, physiological and reproductive 
processes of all living organisms. Mild increases in temperature, of less than 1 °C, may 
simply slightly accelerate cellular processes with no overall detrimental effect. Excessive 
temperature increase can be lethal (Raaphorst et ah, 1979; Dewey et ah, 1977; Dewey, 
1994). Moderate temperature increases may arrest or retard cell division (Mazza et al., 
2004). The effects of a moderate rise above normal physiological temperature can have 
important consequences for developing embryos or foetuses, particularly if the central 
nervous system is involved. The actively dividing cells of the embryonic and foetal central 
nervous system are known to be highly susceptible to changes in temperature (Edwards, 
1969b; Webster and Edwards, 1984; Shiota, 1982, 1988). Interference with neural tissue is 
likely to have significant consequences on growth and development. 



Elevated 
temperatures 
may have 
deleterious 
effects on 
foetal and 
embryonic 
cells and 
tissues 



Medical ultrasonography continues to enjoy increasingly widespread use as an effective 
diagnostic clinical tool. Improvements in resolution and image quality have been 
particularly valuable in obstetrics. Pulsed Doppler (PD) spectral flow analysis and Doppler 
colour flow imaging (CFI) techniques offer the potential to increase diagnostic effectiveness 
and may prove to be valuable diagnostic tools in early pregnancy. There is a growing trend 
to apply new applications of ultrasound, including Doppler, at earlier stages in pregnancy, 
although benefits have not necessarily been demonstrated. Particularly noteworthy is the 
recent movement towards use of spectral Doppler to measure foetal heart rate in the first 
trimester. Owing to the small size of the target, practitioners unfamiliar with the technique 
may require unnecessarily long exposures to obtain the same data that may be had using 
older procedures requiring lower outputs. Furthermore, the introduction of novel imaging 
modalities in ever more sophisticated ultrasound equipment can be accompanied by 
substantial increases in acoustic output (Martin, 2010). This has important consequences 
for the risk of producing thermal bio-effects as the amount of heat deposited in biological 
tissue during ultrasonic examinations is directly related to intensity. 

Diagnostic ultrasound is assumed to be safe. This opinion is supported by a general lack 
of independently confirmed adverse effects from ultrasound exposure in mothers or 
children. However, these conclusions have limited relevance to the way ultrasound is 
currently used in medicine. For example, there are no epidemiological data relevant to 
the increased acoustic outputs available with modern equipment (see Chapters 3 and 9). 
The acoustic outputs of ultrasound scanners in clinical use have increased substantially 
in recent years and are now capable of producing significant heating effects in some 
applications (WFUMB, 1992, 1998). A report by Henderson et al. (1995) described an 
increase by a factor of five in the spatial-peak temporal-average intensity (I spta ) from 
B-mode diagnostic equipment in clinical use in the UK during the period 1991-1995. 
The total acoustic power output in PD mode doubled during that period. More recent 

47 



4 Ultrasound-induced heating and its biological consequences 



Increasing 
acoustic 
output 
levels mean 
that there 
is a greater 
possibility of 
biologically 
significant 
heating 



publications (Duck and Henderson, 1998; Henderson et al., 1997; Whittingham, 2000; 
Martin, 2010) show that the trend for increasing acoustic output has continued to current 
levels that are capable of producing biologically significant heating. 

While maximum outputs have increased, the only effective regulatory body, the Food 
and Drug Administration (FDA), has relaxed the upper limit on intensity that can be 
applied in obstetric ultrasonography in the USA. Under this new scheme (AIUM/NEMA, 
1992) equipment that incorporates an output display may deliver acoustic intensity to the 
embryo or foetus that is almost eight times higher than equipment regulated under the 
Previous application-specific scheme. The rationale for this change is that the responsibility 
is placed on the ultrasound diagnostician to make risk/benefit assessments, based on 
information provided by the equipment output display, and to decide on the appropriate 
examination exposure conditions for each operating condition. The effectiveness of such 
risk assessment depends on the accuracy of the information given in the output display 
and on the ability of the diagnostician to understand it. 



The American Institute of Ultrasound in Medicine and the National Electrical 
Manufacturers Association (AIUM/NEMA, 1992) developed an Output Display Standard 
that includes the thermal index (TI) as an approximation of temperature increase on which 
to assess the potential for thermally mediated biological effects. The TI is calculated from 
the source acoustic power divided by the power needed to raise tissue temperature by 
1 °C. Whilst there is obvious merit in a system that encourages the end-user to become 
aware of safety issues, the TI has some limitations: (a) it does not consider the effect of 
dwell time, i.e. duration of temperature increase at a single point in tissue; and (b) it is 
virtually impossible to predict maximum temperature increase in heterogeneous tissue 
with accuracy. It has been shown that the TI can underestimate the actual temperature 
increase in tissue and that manufacturer estimates of intensity, on which the TI is based, 
can also be significantly in error (Jago et al., 1995). In an assessment of TI for Doppler 
equipment Shaw et al. (1998) concluded "The values of TIS and TIB displayed on the 
scanner should not be taken as the absolute maximum possible temperature rise. The 
worst-case temperature rise may be three times higher than the displayed value." 



The amount 
of heating 
that may be 
expected is 
related to the 
absorption 
coefficient 

The absorption 
coefficient 
increases as 
frequency 
increases 



4.2 Mechanism of ultrasound heating 

During clinical ultrasonographic examinations, an ultrasound beam is transmitted into 
biological tissue. Some of the incident energy is reflected back from interfaces between 
biological tissues to produce echographic images while some is absorbed and converted to 
heat (see Chapter 1). The amount of heat generated depends on the type of examination, the 
acoustic output and the tissue properties. In particular, heating is mostly dependent upon 
the ability of the tissue to absorb, rather than reflect or disperse, ultrasonic energy. For each 
tissue, this ability is quantified by its acoustic absorption coefficient. Generally, more dense 
materials such as bone and teeth have higher absorption coefficients than less dense tissues 
such as liver or muscle and are, therefore, heated to a greater extent than soft tissue. The 
absorption coefficient of bone may be as much as 50 times greater than most soft tissue. For 
example, the average value for brain tissue is 0.2 dB cm 1 MHz 1 while that for mineralized 
bone is lOdBcirr 1 MHz 1 Puck, 1990). As is obvious from the form of the absorption 

48 



Ultrasound-induced heating and its biological consequences 4 



coefficient, the quantity of thermal energy deposited in tissue by an acoustic wave increases 
with the frequency of the wave. Thus the higher frequencies used for imaging superficial 
structures will produce a greater temperature rise in a shorter time than will the lower 
frequencies needed to penetrate to, and image, regions deeper into the body. 



Another important factor for thermal bio-effects is the rate of ultrasound-induced 
heating. The rate of heat deposition in bone is an order of magnitude faster than in soft 
tissue. Therefore, from a safety perspective, the tissue that has the greatest potential for 
bio-effects from ultrasound-induced heating is bone, or developing bone. The extent of 
risk depends on the acoustic exposure conditions and the sensitivity of the target. Tissues 
lying close to, or in contact with, bone are also at risk of heating by conduction from the 
bone. Note that it is particularly important to minimize eye exposures in the foetus and 
adult due to the relatively low perfusion in the eye, particularly in the lens, which thus 
has reduced capability for heat dissipation. As actively dividing cells are most susceptible 
to damage by heat, the foetal cerebral cortex, situated close to the skull bone, is at risk of 
damage by ultrasound-induced temperature increase. 



The extent of 
thermal risk 
depends on the 
heat sensitivity 
of the target 



The volume of heated tissue is determined by the ultrasound beam dimensions. 
For transcranial insonations the transducer can act as a source of heat and may be an 
important factor contributing to brain heating (see Chapter 1). The amount of tissue 
heating that is achieved is limited by the dissipating effects of conduction and convection. 
Blood flow plays an important role, such that highly vascular organs such as the liver 
or kidney are less affected by heating than bone, which has relatively poorly developed 
vasculature. Narrow focused beams that are used in PD and M-mode applications have 
a large temperature gradient between the centre of the beam and the surrounding tissue, 
so that heat is rapidly dissipated by conduction. This is especially true at the focus of 
static beams. In this case vascular perfusion has little additional cooling effect, i.e. there is 
minimal reduction of the amount of ultrasound-induced heating (NCRP, 1992). 



4.3 Temperature increase from diagnostic ultrasound 

The risk of thermal bio-effects in obstetric ultrasound is assessed from the level of 
heating in susceptible tissues, such as in the central nervous system. As the greatest 
temperature increase occurs when bone is situated within the ultrasound beam the 
temperature elevation induced in obstetric exposures depends on the foetal gestational 
age; bone becomes denser and thicker with advancing foetal development. Using the 
mouse skull as a model for human foetal insonation, Carstensen et al. (1990) recorded 
temperature elevations greater than 5 °C after 90 s in anaesthetized animals exposed to 
ultrasound at an intensity (f ta ) of 1.5 W cm" 2 . The -6 dB focal beam width was 2.75 mm. 
The greatest temperature increase was observed in older mice where the rate of rise 
was such that a 4°C elevation above the baseline temperature was achieved within 15 s 
of exposure. The maximum ultrasound-induced temperature increase measured after 
death was approximately 10% higher indicating that blood perfusion in the living animal 
provided a modest cooling effect to counteract the heating. Horder et al. (1998a) showed 
negligible difference between the live and post mortem maximum measured temperature 
rise at the skull/brain interface in guinea-pig foetuses insonated in utero at 57-61 days of 

49 



The sites of 
greatest potential 
heating are 
bone surfaces. 
In the foetus, 
the amount of 
heating to be 
expected rises 
with increasing 
calcification 



4 Ultrasound-induced heating and its biological consequences 



gestational age (dga). A mean maximum temperature increase of 4.9 °C was measured at 
the inner aspect of the skull parietal bone after 120 s exposure to I. ta intensity 2.5 W cm 2 . 
The mean maximum temperature increase was 2 °C at 5 mm depth in the brain. 



Heat loss from 
perfusion is less 
important for 
narrow beams 



In older foetuses insonated in utero near to term (62-67 dga) the mean maximum 
temperature increase was reduced by approximately 12% to 4.3 °C as a result of cooling 
from the more substantially developed cerebral vasculature (Horder et ai, 1998b). These 
studies independently confirmed that significant temperature increase can be produced in 
bone and nearby tissue, and that blood flow provided minimal cooling effects in narrow 
focussed ultrasound beams such as those used for diagnostic scanning, including obstetric 
applications. The later studies measured the temperature increase on the inside of the foetal 
skull adjacent to the cerebral cortex at a biologically significant site. The acoustic exposure 
conditions were higher than typical values but were, nevertheless, within the published 
range of outputs of ultrasonographic equipment (Duck and Henderson, 1998; Martin, 2010). 



The transducer 
is an important 
secondary source 
of heat 



The transducer provides a substantial source of heating, by conduction, in soft tissue 
examinations (WFUMB, 1992). This is particularly important for pulsed transducers, 
which are inefficient in converting electrical to acoustic energy. Heating is localized close 
to the transducer. This should be taken into consideration for transcranial examinations 
where the transducer directly heats bone. There are also implications for safety for 
intracavitary applications, particularly where there is a trend towards increased power 
outputs in gynaecological examinations using the endovaginal route. There is a potential 
risk of inadvertently exposing an unknown pregnancy to heat. 



Whole-body 
hyperthermia 
gives slow rates 
of heating 



4.4 Effects of ultrasound-induced heating 

Estimates of the risk of thermally mediated adverse bio-effects of ultrasound are based on 
data from whole-body hyperthermia studies. However, there may be differences in the 
result of ultrasound-induced heating. When ultrasound energy impinges on biological 
tissue it results in a temperature rise that is rapid compared to the tens of minutes 
required to elevate foetal temperature under whole-body heating conditions. Studies of 
external whole-body heating of rats have reported time constants on the order of 13min 
per 1°C elevation in core temperature (Kimmel et ah, 1993). Conventional sources of 
heating warm the surface of the body, allowing the thermoregulatory system to control 
core temperature via heat receptors in the skin. Such relatively slow heating can also 
trigger the synthesis of specific heat-shock proteins that may afford some degree of 
thermotolerance to dividing cells (Walsh et ah, 1987). However, as this process takes up to 
15 min there is insufficient time for it to occur as a result of ultrasound-induced heating. 



Significant heating 
by ultrasound 
requires a 
stationary beam 



In imaging procedures using scanned beams any single point tissue target is 
interrogated for only fractions of a second each time the beam sweeps past so that 
there is little opportunity to heat a specific tissue target. To achieve significant amounts 
of heating requires that the ultrasound beam be fixed in relation to a tissue target so 
that all the energy in the beam is directed onto that target. This occurs in PD spectral 
measurement techniques and in some CFI applications, such as, for example, studies 
of foetal breathing. 

50 



Ultrasound-induced heating and its biological consequences 4 



The rapid onset of ultrasound-induced heating and its related bio-effects was demonstrated 
in a study showing abnormalities in proliferating bone marrow cells in adult guinea 
pigs. An original study reported heating to 42.5-43.5 °C (3-4 °C above normal guinea-pig 
temperature) in a hot-air incubator for 60 min produced bizarre multisegmented nuclear 
abnormalities in neutrophils extracted from the femurs (Edwards and Penny, 1985). 
The bone marrow is the main site for blood formation in the third trimester. When the 
femur was exposed to ultrasound to elevate the temperature in the bone marrow, the 
same abnormal nuclear division in neutrophils was produced when ultrasound exposure 
elevated the temperature by 3.5 °C above normal for 4 min (Barnett et al., 1991). This 
damage threshold is remarkably similar to that for brain anomalies reported in studies 
using water immersion heating in different animal species, i.e. exposure to a temperature 
increase of 4 °C for 5 min. 



Abnormal 
division of 
neutrophils 
has been 
observed 
following 
heating by 
ultrasound 



The normal homeostatic processes regulate body temperature by increasing the heart rate 
and blood flow to heated regions of the body. In postnatal life this is carried out through 
the hypothalamus, which responds to change in the temperature of blood flowing through 
it. In a study designed to test whether the foetus is able to detect and respond to localized 
mild temperature increase, Horder et al. (1998c) insonated the hypothalamus region of 
the brain of guinea-pig foetuses in utero and measured temperature and foetal heart rate. 
Insonations for 120 s that produced a temperature increase of 1.5 °C at the sphenoid bone 
adjacent to the hypothalamus did not elicit a change in the foetal heart rate. Either the 
foetus is unable to react to localized temperature rise or the extent of temperature increase 
was not biologically significant. If the latter inference is correct, it is further evidence that 
a temperature increase of 1.5 °C above normal body temperature does not present a risk 
to the embryo or foetus. This is consistent with a recommendation of the WFUMB that 
"A diagnostic exposure that produces a maximum in situ temperature rise of no more 
than 1.5 °C above normal physiological levels (37 °C) may be used clinically without 
reservation on thermal grounds" (WFUMB, 1998). 



There is evidence from sensitive studies using embryo culture systems that the effects 
of pulsed ultrasound may be enhanced by a moderate temperature increase (Angles 
et al., 1990; Barnett et al., 1990). Exposure to ultrasound at a spatial peak, temporal- 
average intensity (J spta ) of 1.2 W cm -2 for 15 min at normal temperature produced no 
adverse developmental effects. However, when ultrasound was applied together 
with a modest temperature increase of 1.5 °C (absolute temperature 40 °C) there was a 
significant retardation of growth and reduction of head:body size. As there are few data 
on the biological effects of interaction of ultrasound with tissues that have a pre-existing 
temperature elevation, it is prudent to exercise care when using ultrasound in febrile 
obstetric patients. The published international consensus is that "Care should be taken to 
avoid unnecessary additional embryonic and foetal risk from ultrasound examination of 
febrile patients" (WFUMB, 1998). 



Ultrasound 
scanning of 
febrile obstetric 
patients requires 
particular care 



4.5 Biological effects of hyperthermia 

From the scientific data on biological effects of hyperthermia, it is generally accepted that 
tissues containing a large component of actively dividing cells are particularly sensitive 

51 



4 Ultrasound-induced heating and its biological consequences 



to the effects of heat (Dewey et al., 1977). Abnormalities in cellular physiology and 
biochemical processes can occur following an increase in temperature above normal basal 
levels. The interference with normal rates of enzyme synthesis and reactions can affect 
the way cells grow and divide and may even lead to abnormalities in DNA synthesis and 
repair processes. 



4.5.1 Prenatal animals 



Some stages of 
development 
are more 
sensitive to 
heat damage 
than others 



A large body of scientific data clearly shows that there are critical periods during gestation 
when the embryo and foetus are susceptible to thermal insult (Bell, 1987; Edwards, 1986, 
1993; Edwards et al., 1995; Kimmel et al., 1993). During formation of the embryonic neural 
plate and closure of the neural tube, perturbations can result in severe neural tube defects, 
retarded brain development, exencephaly and microphthalmia in guinea pigs (Edwards, 
1993). Kimmel et al. (1993) observed craniofacial anomalies and other foetal skeletal 
anomalies in rats following maternal exposure to increased temperature. Exposure to heat 
at pre-organogenesis stages in development can result in cardiovascular abnormalities 
(Edwards, 1993). If the embryo is heated at later stages (e.g. day 13 of gestation in rats) 
the skeletal and visceral systems can be affected. A moderate temperature increase (2 °C 
above normal), together with exercise, has been shown to induce a range of teratogenic 
effects in rats (Sasaki et al, 1995). Non-specific effects such as foetal weight reduction are 
also associated with intrauterine heating or maternal stress. A brief list of effects observed 
in animals, the experimental animal studied and the temperature and duration of the 
exposure, is given in Table 4.1 (Church and Miller, 2007). 



Actively 
dividing cells 
are particularly 
sensitive to 
thermal damage 



One conclusion that may be drawn from Table 4.1 is that different tissues may have 
diff erent sensitivity to the heating produced by diagnostic ultrasound (vide Barnett 
et al., 1997). Actively dividing foetal neural tissue is highly sensitive to damage by 
heat (Edwards et al., 1995) while adult tissue is generally more resistant and tolerant. 
If a transient temperature increase arrests mitotic cell division in the brain during 
embryonic development the resulting neural deficit may not be restored, although the 
foetus may continue to develop and appear morphologically normal. Such brain-growth 
retardation is a common result of hyperthermia in animal experiments (Graham et al., 
1998). Developing embryos may mount a protective response to sublethal hyperthermia 
that temporarily arrests the normal process of cell division. This phenomenon has 
been observed in the brains of rodents where normal cell division lapsed for up to 8h 
following heat treatment (Edwards et al., 1974; Upfold et ah, 1989). Meanwhile, heat-shock 
proteins (hsp) may be synthesized at the expense of normal neural proteins. On recovery, 
normal cell division resumed with the foetus appearing morphologically normal, albeit 
smaller and with a substantial neural deficit. In embryonic development a lapse of a few 
hours can lead to substantial delay or disturbance in neurological development. Non- 
deforming retardation of brain growth and reduced learning performance are common 
abnormalities in the offspring of moderately heat exposed pregnant guinea pigs. These 
defects can be caused both during early and later foetal growth (Edwards, 1993). In 
general, embryos are more susceptible to damage than foetuses due to the high rate of 
cellular activity during organogenesis. However, continually developing organ systems 
such as the brain remain susceptible to heat throughout pregnancy. 

52 



Ultrasound-induced heating and its biological consequences 4 



The effects of increased temperature on biological systems have been extensively reviewed 
(Barnett et al, 1994; Edwards, 1993; Miller and Ziskin, 1989; Edwards et al, 2003; Church 
and Miller, 2007; Abramowicz et al, 2008; O'Brien et al, 2008). In their study, Edwards 
et al. (2003) made some crucial observations. (1) The relative timing of a particular 
developmental stage of interest does not scale linearly with the average length of gestation, 
making extrapolation of animal results to human exposures difficult. For example, the 
thermally sensitive stage of neural tube closure occurs at day 8.5 in the mouse, day 9.5 in 
the rat, days 13-14 in guinea pigs, and days 20-28 in humans. Similarly, neurogenesis, 
neural cell migration and the development of bodily structures occurs during days 12-15 
in mice, days 13-18 in rats, days 20-35 in guinea pigs, and days 56-126 in humans. (2) The 
effect of a given heating regimen will depend on the stage of gestation at which it occurs, 
regardless of the species involved. For example, hyperthermia during the period of neural 
tube closure may result in the death of neuronal cells leading to anencephaly, spina bifida 
or encephalocoele, while the same exposure a few days or weeks later (depending on the 
species) during the period of neuronal cell proliferation for brain development may also 



The developing 
foetus is 
particularly 
sensitive to 
hyperthermia 
during the 
period of neural 
tube closure 



Table 4.1. Selected thermally induced teratogenic effects (from Church and Miller, 2007). 



Thermal effect Animal/species Temperature (°C) Duration (min.) Source 



Abortion 


Monkey 


40.6 


72 


Hendrickx 
et at., Ly/y 


Absence of 
optical vesicles 


Rat 


43.0 


7.5 


Walsh et al, 

170/ 


Anencephaly 


Rat 


41.0-43.5 


41) 


Edwards, 1968 


Behavioural 
anomalies 


Marmoset 


41.5 


60 


Poswillo 
et al, 1974 


Cardiac and 

vascular 

anomalies 


Chicken 


41.0 


3180 


Nielson, 1969 


Developmental 
anomalies 


Rat 


42.0 


41) 


Skreb and 
Frank, 1963 


Exencephaly 


Mouse 


42.3 


5 


Webster and 
Edwards, 1984 


Eye defects 


Chicken 


41.0 


1440 


Nielson, 1969 


Limb, toe & tail 
defects 


Rat 


41.0-43.5 


40 


Edwards, 1968 


Micrencephaly 


Guinea pig 


43.0 


60 


Edwards, 
1969fl,b 


Microphthalmia 


Rat 


41.0 


60 


Germain 
et al, 1985 


Posterior 
paralysis 


Mouse 


43.0 


60 


Permycuik, 
1965 


Resorption 


Guinea pig 


42.9 


60 


Edwards, 1967 


Scoliosis 


Monkey 


40.6 


72 


Hendrickx 
et al, 1979 


Skeletal defects 


Marmoset 


41.5 


60 


Poswillo 
et al, 1974 


Tooth defects 


Rat 


38.9 


720 


Kreshover and 
Clough, 1953 



53 



4 Ultrasound-induced heating and its biological consequences 



kill cells, but the outcome may be micrencephaly (small brain) rather than anencephaly 
(no or very rudimentary brain), or simply learning disorders not accompanied by obvious 
structural abnormalities. The interested reader should consult Edwards et al. (2003) and 
Abramowicz et al. (2008) for more complete discussion of teratogenesis in experimental 
animals and its relation to potential effects in humans. 



4.5.2 Prenatal humans 



Hyperthermia 
is a suspected 
teratogen in 
humans 



Although relatively few studies on the effects of heating during gestation in humans 
have been published, many of the abnormalities reported in animal studies have also 
been observed in humans following in utero febrile episodes. For example, Smith et al. 
(1978) found that maternal febrile illness producing a body temperature above 38.9 °C 
in early stages of pregnancy was associated with development of foetal anomalies. In a 
confirmatory study, Chambers et al. (1998) reported that the offspring of women who 
had a "high" fever (T > 38.9 °C for at least 24 h) during pregnancy had a increased rate of 
major malformations, 15.8% compared to 4.5% among controls, as well as an increased 
incidence of minor malformation. Although none of the differences observed were 
statistically significant, the authors conclude that high maternal fever early in pregnancy 
is a human teratogen. Similarly, Little et al. (1991) demonstrated a statistically significant 
association between a maternal fever of 38.3 °C for 24 h or more and subsequent 
abdominal wall defects in their offspring. Retrospective studies indicate that mothers of 
babies with various malformations of the central nervous system experienced increased 
prevalence of febrile illness during early pregnancy (Layde et al., 1980; Pleet et al., 1981; 
Shiota, 1982; Spraggett and Fraser, 1982). For example, Pleet et al. (1981) observed a 
relationship between maternal hyperthermia during weeks 4-14 of gestation and the 
induction of defects of the central nervous system (mental deficiency), altered muscle 
tone (hypotonia) and facial anomalies such as microphthalmia or cleft lip. Fraser and 
Skelton (1978) also found an association between microphthalmia and fever during the 
first trimester. Epidemiological studies by Erickson (1991) and Tikkanen and Heinonen 
(1991) have shown a relationship between maternal hyperthermia and congenital heart 
defects. Graham et al. (1998) provide a summary of the relationship between maternal 
hyperthermia and congenital birth defects. Based on the similarity of responses across 
species, hyperthermia is strongly suspected of being a teratogen in humans (Edwards, 
1986, 2006; Edwards et al, 1995; Shepard, 1982, 1989; Suarez et al, 2004; Moretti et al, 
2005). 



One report that has had a profound impact on our perception of the potential 
consequences of foetal hyperthermia is the comprehensive study on the effects of 
hyperthermia in animals and the potential for exposure to diagnostic ultrasound to 
induce adverse biological effects in the foetus that was published by Miller and Ziskin 
(1989). Two of their conclusions warrant special att antion here: (1) hyperthermic 
treatment during gestation resulted in teratogenic effects whose magnitude varied 
with exposure temperature and duration such that the higher the temperature of the 
exposure medium, the shorter the time needed to cause an effect; and (2) on a plot 
of exposure duration versus exposure temperature, a convenient boundary could be 
drawn below which there were no observed effects. The plot referred to in item 2 is 

54 



Ultrasound-induced heating and its biological consequences 4 



reproduced here as Figure 4.1. Each point represents either the lowest temperature 
reported for any duration or the shortest duration for any temperature reported for a 
given effect. A solid line connecting points indicates multiple data points relating to a 
single study or effect. The dashed line defines a lower boundary (f =f =1, see below) 
for observed biologic effects. This analysis was widely read and accepted by the scientific 
(AIUM/NEMA, 1992; NCRP, 1992, 2002; WFUMB, 1992, 1998; BMUS, 2000; Herman 
and Harris, 2002) and regulatory (FDA, 1997; IEC, 2007) communities. It provided 
the rationale for concluding that for obstetric diagnostic ultrasound examinations, 
a temperature elevation of 1°C or 1.5 °C could be applied for any duration without 
concern for a thermal bio-effect. 



However, two significant features of the data underlying Figure 4.1 are easily missed. 
First, every experiment involved whole-body heating of the maternal animal, generally 
in either a heated water bath or in a hot-air incubator. This is important because no 
foetus is heated to the experimental exposure temperature instantaneously. There is 
always some time lag as the mother's body attempts to compensate for the sudden 
increase in environmental temperature, so the actual time that the foetus experiences the 
exposure temperature used in Figure 4.1 will be less than the duration of exposure. 
Second, different experimental animals have different normal physiological 
temperatures. This is important because we do not know whether a 2 °C rise in an 
animal having a normal temperature of 39 °C, e.g. the 



y 45 - 



0J 



I 

(2 

B 
o 

X 

m 




10 



100 



1000 



Exposure Duration (min) 

Figure 4.1. Thermally induced biological effects, listed in Table 4 of Miller and Ziskin (1989), that 
have been reported in the literature and for which the temperature elevation and exposure 
duration are provided. The dashed line defines a lower boundary (r 43 = 1) for observed biologic 
effects. The open square and arrowed "Edwards" indicate the exposure temperature and 
duration (60 min, 43 °C) for the Edwards (1969b) hyperthermia treatment of pregnant guinea 
pigs; figure taken from Church and Miller (2007). 

55 



4 Ultrasound-induced heating and its biological consequences 

guinea pig, will have the same effect in an animal with a normal temperature of 37 °C such 
as the human being. In other words, which is the critical factor, the absolute temperature 
or the change in temperature? Opinions vary, but consider an extreme example: the pigeon 
lives its entire life at 43 °C (Edwards, 1986), a temperature that would be fatal to a human 
in a matter of days. Therefore, until the question is resolved, prudence dictates that the 
change in temperature be given greater weight. 

4.5.3 Postnatal individuals 

Most of the preceding discussion has focused on the effects of heating on the embryo 
or foetus. This is entirely appropriate because prenatal subjects are more sensitive 
to a wide range of external stimuli than the same individuals would be if the same 
stimulus occurred later in life. In addition, if a particular ultrasound exposure produces 
harm in a child or adult, that person can respond in ways that are obvious to the 
sonographer or physician, but the foetus lacks similar mechanisms for providing direct 
feedback to the sonographer. While relatively few reports exist relating diagnostic 
ultrasound exposure conditions to non-foetal thermal effects, a recent report by O'Brien 
et al. (2008) p rovides a valuable review of the effe cts of hyperthermia in postnatal subjects. 

O'Brien et al. (2008) compiled data from a number of sources for temperature thresholds 
for thermal damage to non-foetal tissue produced by single exposure durations as short 
as 0.1s, see their Table 2. Using these data, they were able to construct a boundary below 
which no thermal damage is observed in non-foetal soft tissue. The data and boundary 
are shown in Figure 4.2; see O'Brien et al. (2008) for details. The report suggests that for 
non-foetal soft tissue and for scanning conditions consistent with conventional B-mode 
ultrasound examinations for which the exposure durations at the same in situ locations 
would be less than a few seconds, the allowable maximum temperature increase could 
be relaxed relative to values represented by the conservative boundary line for longer 
exposures (O'Brien et al, 2008). 

4.6 Quantification of heating — thermal dose 

There is a well-known relationship between an elevation in temperature above the 
normal physiological level and the time needed to induce a deleterious effect in a 
biological system (NCRP, 1992, 2002; Edwards et al, 2003; Miller and Ziskin, 1989). 
Simply put, the higher the temperature rise, the shorter the time needed to produce the 
eff act. Dewey et al. (1977)and Sapareto and Dewey (1984)s howed that for two exposures 
at two different temperatures, the ratio of the minimum time t required for the lower 
temperature T 1 to produce an effect, to the time f, required for the higher temperature 
T 2 to produce the same effect increased by a constant multiple for each degree of 
temperature difference. For example, if the temperature difference T.-T =1°C, then 
tJt 2 = R, while if T -T =2°C, then f 1 /f 2 =R 2 , etc., where R is the thermal normalization 
constant. In general, tJt=R Tr ~ T ' . The time used to quantify thermal exposures is termed 
the "thermal dose" (f REF ): 

W = « WT (4.1) 



Thermal dose is 
used to quantify 
exposure to heat 
and compare 

thermal effects 



56 



Ultrasound-induced heating and its biological consequences 4 




0.1 1 10 100 1000 

Exposure Duration (s) 



Figure 4.2. Temperature thresholds for damage to non-foetal tissues from single exposures to 
heat. The solid line represents a conservative boundary for exposure durations in non-foetal 
soft tissue. See O'Brien et al. (2008) for details. 



Empirical values of R vary among species, tissues and biological endpoints. They also are 
temperature-dependent, with R~2 for T>43 °C, increasing by a factor of 2-3 for T<43 °C. 
For simplicity, the values for R are usually fixed at R = 2 for T>43 °C and R = 4 for T<43 °C 
(Sapareto and Dewey, 1984; Dewey, 1994). 

Miller et al. (2002) salved the problem of normalizing experimental results obtained 
in animal species having different core temperatures by subtracting the animal's core 
temperature T from both T REF and T. In this way, Equation 4.1 may be written: 

t REF = fR Ar «-- AT (4.2) 

where AT REF = (T REF - T) and AT=(T-T). The numerical values of Equations 4.1 and 4.2 are 
the same, but rather than a fixed reference temperature, the reference is now written in 
terms of the increase above the animal's core temperature. The effective temperature for 
a teratological effect simply scales from the animal's normal physiological temperature 
(Dewey et al, 1977; Raaphorst et ah, 1979; Sarge et al., 1995). The transition point for R is 
assumed to be at a temperature elevation of 6-7 °C. The value of Equation 4.2 is that it 
normalizes data to a common denominator, the organism's normal temperature and, thus, 
allows comparison of data among organisms having different core temperatures (Church 
and Miller, 2007). 

Equation 4.2 can be applied to the data depicted in Figure 4.1 and thus normalize them 
to the increases above each animal's normal physiological temperature. The result of 



57 



4 Ultrasound-induced heating and its biological consequences 



A temperature 
rise of 3.5 °C for 
1 min represents 
a "safe" thermal 
dose to the 
foetus 



normalization is shown in Figure 4.3. Each datum was derived from an analysis of 
a heating-cooling profit of a pregnant laboratory animal (rat, mouse or guinea pig) 
which yielded foetuses with teratologic anomalies (Miller et al., 2002). The original "safe" 
boundary line in Figure 4.1, i.e. t=l min, is now given by f 43 37 = t 6 = l min, shown by the 
dashed line. About half of the data points in Figure 4.3 are below this line, showing that 
temperature exposures previously considered safe because they fell below the boundary 
in Figure 4.1, i.e. below t=l min, are actually potentially harmful. A new "safe" boundary 
might be drawn at f 4 =lmin if it were not for a single point derived from the data of 
Kimmel et al. (1993). While the data collected by Kimmel et al. (1993) include seemingly 
minor anomalies and also those with high natural background rates, prudence suggests 
that the "safe" line be drawn at f = 1 min in Figure 4.3, as shown by the dotted line; this is 
equivalent to f 4 0 = 0.5 min. 



4.7 Thresholds for biological effects 

Most of the early studies on whole-body heating were not designed a priori to identify 
threshold levels but, instead, set out to demonstrate that severe developmental 
abnormalities can be produced by heat exposure. In most cases there was no record of 



o 

o 



I 4 

'+-> 
(0 

> 

a) 3 

LU 

<b 

3 2 
2 

I 1 

0) 



i i i i i i Q- 



o 

oo d&sTo o 



-i — i— n 



i i i 1 1 1 



Edwards 



© CD OO OCN^ 



o odbo^ 

o \ 



Kimmel 



10 



100 



1000 



Time at Equivalent AT (min) 

Figure 4.3. Thermal-equivalent core temperature elevations versus time. The dashed line 
is the equivalent lower boundary (t 43 =1) shown in Figure 4.1, and the dotted line is the 
lower boundary for the "transformed" data. The open triangle shows the lowest positive- 
result datum, from Kimmel et al. (1993), Ar = 1.8°C. The open squares and arrows show the 
"movement" of the Edwards' data point from its original location at (60min, 43°C) in Figure 
4.1 to its "final" location at (8.9 min., 3.5 °C). Adapted from Church and Miller (2007). 



58 



Ultrasound-induced heating and its biological consequences 4 

either the temperature elevation or the duration for which it was maintained within the 
foetus. The heat dose was estimated from the maternal core temperature measured per 
rectum. 

4.7.1 Prenatal individuals 

Useful threshold data have been derived from the study by Germain et al. (1985) which 
used water immersion as a means of rapid body heating and detected encephalocoeles 
as the adverse developmental outcome. The normal resting temperature for 50 pregnant 
rats was measured to be 38.5 ± 0.5 °C during daylight (i.e. when less active, resting). The 
shortest exposure that produced abnormalities was lmin at a temperature of 43.5 °C 
(i.e. 5°C increase above the normal physiological temperature for pregnant rats). The 
same brain abnormality was observed after 5min exposure to a temperature elevation 
of 4 °C. A teratogenic threshold of 4 °C increase in temperature maintained for 5 min was 
also observed by Sasaki et al. (1995) in which hyperthermia was achieved in pregnant 
rats made to swim in a water bath. The majority of brain malformations involved 
microphthalmia and encephalocoeles. The resting core temperature measured in all rats 
prior to heat treatment was between 38 °C and 39 °C. Other studies have also reported 
brain anomalies following prenatal exposure to elevated temperature. For example, 
exencephaly was produced in mice (Webster and Edwards, 1984) following intrauterine 
exposure to 42.3 °C (equivalent to a 4.3 °C increase above normal body temperature), for 
5 min. A similar result was reported by Shiota (1988) for exencephaly in mice exposed to 
a 4.5 °C increase above normal body temperature for 5 min. Exposure to a whole-body 
temperature increase of 3.5 °C above normal for 10 min produced exencephaly in mice 
(Shiota, 1988) and microphthalmia in rats (Edwards et al, 1995). 

It is instructive to consider the data from Figure 4.3, for which foetal temperature-time 
profiles are available, in terms of a dose-response relationship. The percentages of 
foetuses exhibiting major craniofacial anomalies in the reports by Germain et al. (1985) 
and Shiota (1988) are plotted as a function of the thermal dose in Figure 4.4. The best- 
fitting line through the data intersects the x-axis at approximately 0.5 min, indicating that 
the data are not inconsistent with the concept of a threshold. Since there is considerable 
scatter in the data, the exact value of any threshold cannot be determined with much 
accuracy. The lower 95% confidence interval on the fit intersects the x-axis at about 
4.2 min, suggesting that the threshold is no greater than this value. 

Naturally one must be cautious when applying these results to the human population. 
Tens of millions of newborns are scanned in utero each year, and if even a small fraction of 
ultrasound examinations result in hyperthermia-induced anomalies, this would represent 
a large number of affected individuals (Miller et al., 2002). A similar situation exists 
when analysing dose-response data and estimating probability of effects at low doses 
of ionizing radiation, in spite of the fact that many more studies have been conducted in 
that area. With this in mind, the most recent report on the Biological Effects of Ionizing 
Radiation (BEIR VII; NA/NRC, 2006) adopts a linear no-threshold model for estimating 
risk from low-level exposures. 



A temperature 
rise of 4.0 °C 
for 0.5 min 
may represent 
a threshold for 
teratological 
effects 



59 



4 Ultrasound-induced heating and its biological consequences 



4.7.2 Postnatal individuals 



Non-foetal soft 
tissue is at least 
two orders of 
magnitude less 
sensitive to heat 
than foetal tissue 



The thermal exposures required to damage a range of non-foetal tissues have been reported 
in terms of thermal dose (see below) for a variety of animals and humans (Dewey, 1994; 
Lyng et ah, 1991). These doses are given in Table 4.2 for t & Since much of the data were 
obtained from animals having normal physiological temperatures different from that of 
the human, the doses are also given in terms of a 4 °C temperature rise, I,. It is significant 
that even the smallest value for t v 80 min for damage to the kidney, is more than 2 orders 
of magnitude greater than the "safe" value determined from Figure 4.4 for the induction 
of major developmental anomalies in the foetus. 



4.8 Thermal index 

In order to determine the effects of a particular hyperthermic exposure on a particular 
tissue or organ, it is necessary to know the temperature-time profile of exposure and also 
to be able to compare a known level of effect produced in the same or similar tissue by a 
known hyperthermic exposure. This is obviously well beyond what might be expected 
of any practicing physician or sonographer, no matter how knowledgeable they may be. 
To provide some level of guidance to users of diagnostic ultrasound, a joint committee 




0 5 10 15 20 25 30 35 



Thermal Dose, q (min) 

Figure 4.4. Percentage of foetal defects versus r 40 based on the results of Germain (G) et a/. 
(1985) in rats, and Shiota (S) (1988) in mice. The author-specific designations refer to the 
maximum temperature increase (AT) above the measured normal physiological temperature. 
The solid line is the best fit to the data, %Defects = 3.893 t = 1.852, r 2 = 0.604, and the dashed 
curves give the 95% confidence interval around the fit. Adapted from Miller et a/. (2002). 

60 



Ultrasound-induced heating and its biological consequences 4 



Table 4.2. Thermal dose values, r 43 and f 4 , for selected tissues (adapted from O'Brien ef. a/., 
2008). 



Ticci i o 


A n irma l/cnorioc 
/-\illlllal/:>|Jtr*_ltr:> 


Mean core 
temperature (°C) 




' 4 limn; 


Muscle, fat 


Pig 


38.5 


240 


480 


Skin 


Human, rat, 


37, 38, 38 


210 


3360, 840, 840 




mouse 




Oesophagus 


Pig 


38.5 


120 


240 


Cartilage 


Rat, mouse 


38, 38 


120 


480 


Breast 


Human 


37 


100 


1600 


Bladder 


Dog, rabbit 


38.5, 39 


on 


1 /'n on 

160, 80 


Small intestine 


Rat, mouse 


38,38 


40 


160, 160 


Colon 


Pig, rabbit 


38.5, 39 


30 


60, 30 


Liver 


Dog, rabbit 


38.5, 39 


30 


60, 30 


Brain 


Cat, dog 


39, 38.5 


25 


25, 50 


Kidney 


Mouse 


38 


20 


80 



of the AITJM, the NEMA, and the FDA was formed to develop an on-screen display to 
maintain and enhance patient safety. Their work resulted in the creation of two general 
indices, the mechanical index (MI) and the TI, to provide information to the user on the 
output level of their machine and how a change in output would affect the likelihood of 
inducing a biological effect in the patient. These indices are now part of the regulations 
governing the manufacture, sale and use of diagnostic equipment in much of the world 
(FDA, 1997; IEC, 2007, 2010). The MI is discussed in Chapter 5. The TI is discussed in 
Chapter 2 and below. 

The basic definition of the TI is: 

TI = (4.3) 

W, 

"deg 

where W g is the is the source power of the diagnostic ultrasound system, and W deg is the 
source power required to increase the temperature of a specific tissue model by 1°C. 
The model assumes a very low level of attenuation (O.SdBcm^MHz -1 ), thus it will 
overestimate the temperature rise expected in real tissue. The TI is conservative in this 
regard. There are three different categories of TI corresponding to different combinations 
of soft tissue and bone that are commonly encountered while imaging patients. Each 
category uses one or more models based on system information, including transducer 
aperture, acoustic beam dimensions and imaging mode. The categories are TIS for soft- 
tissue imaging, TIB for imaging in a non-scanning mode when bone is at or near the 
focus, and TIC for imaging when bone is near the surface, e.g. adult cranial applications 
(Abbott, 1999). 



Two points should be understood in regard to the MI and TI. First, they are not true safety 
indices in that no particular value is directly related to any quantifiable level of risk of 

61 



4 Ultrasound-induced heating and its biological consequences 



The Ml and Tl 
enable application 

of the ALARA 
principle to ensure 

patient safety 



doing harm to the patient. The MI and TI are output indices because their values are related 
to specific output parameters and transducer (probe) characteristics. However, it is self- 
evident that output levels are related to risk in some way, even if the connection between 
the two remains poorly understood. Second, the MI and TI are relative rather than absolute 
quantities. In other words, an examination performed at a TI of 2 represents greater risk to 
the patient than the same examination performed at a TI of 1, although the absolute risk is not 
necessarily known. Although this fact may be unsettling to some, it is also not particularly 
important because when properly used, the indices still fulfill their goal of helping to ensure 
patient safety. This is done by application of the ALARA (as low as reasonably achievable) 
principle (NCRP, 1990), which ultrasound borrows from radiation biology. 



A third point may be made in regard to the TI. The TI does not include time as one 
of its input parameters. This simplification is necessary owing to the severe technical 
challenges that must be met and the enormous computational overhead required for a 
diagnostic machine to determine exactly when it began imaging a region of tissue and 
when it has been moved to a different region. Various possible solutions to the problem 
have been proposed (Lubbers et al., 2003; Church, 2007; Ziskin, 2010), but none has yet 
been adopted by any regulatory agency or manufacturer. 



4.9 Guidelines for safe use of diagnostic 
ultrasound 

Even though the TI is available on-screen, the user still faces the problem of knowing 
how long to image a particular tissue or organ. This problem was recently considered by 
the Safety Group of the British Medical Ultrasound Society (ter Haar, 2010; BMUS, 2010) 
and by Nelson et al. (2009). Both groups recommended restricting the acoustic output to 
no more than that actually required to obtain the necessary diagnostic information. The 
BMUS Safety Group provided two levels of their guidelines, one more basic and easily 
remembered (for obstetric examinations only), and a second that is more detailed in terms 
of maximum duration of scanning at each of several values for TI for various, specified 
applications. Nelson et al. (2009) provided only a relatively simply set of guidelines. The 
latter are also, in a few cases, somewhat more conservative in terms of maximum duration 
for a given TI for obstetric examinations, but they are less conservative for examinations 
in adults. The guidelines suggested by the two groups for setting and monitoring acoustic 
output during scanning may be summarized as: 



4.9.1 Prenatal examinations 

1. TI values less than 0.7 [Nelson, <0.5] should be used unless otherwise required, 
particularly in the first trimester. 

2. More generally, TI values less than 0.7 [Nelson, <0.5] likely can be used for scanning 
times on an extended basis. 

3. TI values greater than 0.5 and up to 1 should be limited to scanning times less than 
60 [Nelson, <30] min. 

4. TI values greater than 2.5 should be limited to scanning times less than 1 min. 



62 



Ultrasound-induced heating and its biological consequences 4 



4.9.2 Postnatal examinations 

1. TI values less than 1 [Nelson, <2] likely can be used for scanning times on an extended 
basis. 

2. TI values greater than 1 and up to 1.5 should be limited to scanning times less than 
120 min, for 1.5 < TIB < 2.0 to 60 min, for 2.0 < TIB < 2.5 to 15 min, for 2.5 < TIB < 3.0 to 
4min,for 3.0 < TIB < 4.0 to 1 min and for 4.0 < TIB < 5.0 to 15 s. [Nelson states that for 
thermal indices in the range 2-6 time should be restricted to less than 30 min] . 

3. TI values greater than 5 [Nelson, >6] should be limited to scanning times less than 
5s [Nelson, 1 min]. 

4. For neonatal studies, one should consider further limiting exposure levels and 
durations as developmental processes are continuing. 

It is emphasized that these are only suggested guidelines and not rules to be rigidly 
followed. This is due to the uncertainty regarding the relationship between acoustic 
output and the induction of biological effects. More importantly, it is understood that 
depending on the scanning conditions and diagnostic requirements of the study, these 
levels may be exceeded for limited periods to obtain diagnostically useful information 
and thus to ensure optimal patient care. 



4.10 Conclusions 

Data from animal studies demonstrate that exposure for 0.5 min to a temperature 
increase of 4 °C above the normal body temperature is potentially hazardous to 
embryonic and foetal development; exposures longer than 5 min involve significant 
risk of harm. Temperature increases >4 °C have been measured at or near bone/soft 
tissue interfaces in the brains of animal foetuses during in utero exposure to conditions 
similar to those available with PD systems. The effects of elevated temperature may 
be minimized by keeping the time for which the beam passes through any point in 
tissue as short as possible. The central nervous system of the first-trimester embryo 
is particularly sensitive to damage and/or modification by heating. The absorption 
coefficient of embryonic tissue is lower than that of bone tissue and as the embryo 
contains no ossified bone, it has a lower risk of suffering bio-effects from ultrasound- 
induced heating. Adult tissue is less susceptible to damage from ultrasound-induced 
temperature rise. 

There are few data on the biological effects of interaction of ultrasound with tissues 
that have a pre-existing temperature elevation. The results of specialized studies using 
rat embryo culture techniques imply that ultrasound-induced biological effects can be 
potentiated by an existing elevated core temperature; however, uncertainties remain about 
the possibility of synergistic effects from such ultrasound interactions. Epidemiology 
data provide little evidence of adverse effects following prenatal exposure to diagnostic 
ultrasound (see Chapter 9). However, there are no data on obstetric exposures with 
modern ultrasonographic equipment operating with a real-time output display, with 
which significant temperature increase can be produced. 



Output guidelines 
may be exceeded 
for a limited time 
if necessary 
to obtain 
diagnostically 
valuable 
information 



63 



4 Ultrasound-induced heating and its biological consequences 



References 

Abbott JG. 1999. Rationale and derivation of the MI and H: a review. Ultrasound Med Biol, 
25, 431-441. 

Abramowicz JS, Barnett SB, Duck FA, Edmonds PD, Hynynen KH, Ziskin MC. 2008. Fetal 
thermal effects of diagnostic ultrasound. / Ultrasound Med, 27, 541-559. 
AIUM/NEMA. 1992. Standard for Real-time Display of Thermal and Mechanical Acoustic 
Output Indices on Diagnostic Ultrasound Equipment. Rockville, MD: American Institute 
of Ultrasound in Medicine. 

Angles JM, Walsh DA, Li K, Barnett SB, Edwards MJ. 1990. Effects of pulsed ultrasound 
and temperature on the development of rat embryos in culture. Teratology, 42, 285-293. 
Asakura N. 2004. Fetal and neonatal thermoregulation. / Nippon Med Sch, 71, 360-370. 
Barnett SB, Walsh DA, Angles JA. 1990. Novel approach to evaluate the interaction of 
pulsed ultrasound with embryonic development. Ultrasonics, 28, 166-170. 
Barnett SB, Edwards MJ, Martin P. 1991. Pulsed ultrasound induces temperature 
elevation and nuclear abnormalities in bone marrow cells of guinea-pig femurs. In 
Proceedings of the 6th WFUMB Congress in Ultrasound, Copenhagen, Denmark, 
#3405. 

Barnett SB, ter Haar GR, Ziskin MC, Nyborg WL, Maeda K, Bang J. 1994. Current status of 
research on biophysical effects of ultrasound. Ultrasound Med Biol, 20, 205-218. 
Barnett SB, Rott HD, ter Haar GR, Ziskin MC, Maeda K. 1997. The sensitivity of biological 
tissue to ultrasound. Ultrasound Med Biol, 23, 805-812. 

BMUS. 2000. The Safe Use of Ultrasound in Medical Diagnosis. 2nd Edition. London, UK: 
British Medical Ultrasound Society & The British Institute of Radiology. 
BMUS. 2010. Guidelines for the safe use of diagnostic ultrasound equipment. Ultrasound, 
18, 52-59. 

Bell AW. 1987. Consequences of severe heat stress for fetal development. In Heat Stress: 
Physical Exertion and Environment, Hales JRS, Richards DAB (editors). Amsterdam, the 
Netherlands: Elsevier Science Publishers BV. 

Carstensen EL, Child SZ, Norton S, Nyborg WL. 1990. Ultrasonic heating of the skull. 
JAcoust SocAm, 87, 1310-1317. 

Chambers CD, Johnson KA, Dick LM, Felix RJ, Jones KL. 1998. Maternal fever and birth 
outcome: a prospective study. Teratology, 58, 251-257. 

Church CC. 2007. A proposal to clarify the relationship between the thermal index and 
the corresponding risk to the patient. Ultrasound Med Biol, 33, 1489-1494. 
Church CC, Miller MW. 2007. Quantification of risk from fetal exposure to diagnostic 
ultrasound. Prog Biophys Mol Biol, 93, 331-353. 

Dewey WC. 1994. Arrhenius relationships from the molecule and cell to the clinic. Int } 

Hyperthermia, 10, 457-483. 

Dewey WC, Hopwood LE, Sapareto SA, Gerweck LE. 1977. Cellular responses to 
combinations of hyperthermia and radiation. Radiology, 123, 463-474. 
Duck FA. 1990. Physical Properties of Tissue. A Comprehensive Reference Book. London, 
UK: Academic Press. 

Duck FA, Henderson J. 1998. Acoustic output of modern ultrasound equipment: is it 
increasing? In Safety of Diagnostic Ultrasound, Barnett SB, Kossoff G (editors). New York, 
NY: Parthenon Publishing Group, pp. 15-25. 

64 



Ultrasound-induced heating and its biological consequences 4 



Edwards MJ. 1967. Congenital defects in guinea pigs following induced hyperthermia 
during gestation. Arch Pathol, 84, 42-48. 

Edwards MJ. 1968. Congenital malformations in the rat following induced hyperthermia 
during gestation. Teratology, 1, 173-175. 

Edwards MJ. 1969a. Congenital defects in guinea pigs: fetal resorptions, abortions and 
malformations following induced hyperthermia during early gestation. Teratology, 2, 
313-328. 

Edwards MJ. 1969b. Congenital defects in guinea pigs: prenatal retardation of brain 

growth of guinea pigs following hyperthermia during gestation. Teratology, 2, 329-336. 

Edwards MJ. 1986. Hyperthermia as a teratogen: a review of experimental studies and 

their clinical significance. Teratog Carcinog Mutagen, 6, 563-582. 

Edwards MJ. 1993. Hyperthermia and birth defects. Cornell Vet, 83, 1-7. 

Edwards MJ. 2006. Review: hyperthermia and fever during pregnancy. Birth Defects Res A 

Clin Mol Teratol, 76, 507-516. 

Edwards MJ, Penny RHC. 1985. Effects of hyperthermia on the myelograms of adult and 
fetal guinea-pigs. Br ] Radiol, 59, 93-101. 

Edwards MJ, Mulley R, Ring S, Wanner RA. 1974. Mitotic cell death and delay in mitotic 
activity in guinea-pig embryos following brief maternal hyperthermia. / Embryol Exp 
Morphol, 32, 593-602. 

Edwards MJ, Shiota K, Smith MSR, Walsh DA. 1995. Hyperthermia and birth defects. 
Reprod Toxicol, 9, 411-425. 

Edwards MJ, Saunders RD, Shiota K. 2003. Effects of heat on embryos and foetuses. Int 
] Hyperthermia, 19, 295-324. 

Erickson JD. 1991. Risk factors for birth defects: data from the Atlanta birth defects 
case-control study. Teratology, 43, 41-51. 

FDA. 1997. Information for Manufacturers Seeking Marketing Clearance of Diagnostic 
Ultrasound Systems and Transducers. Rockville, MD: Center for Devices and Radiological 
Health, US Food and Drug Administration. 

Fraser FC, Skelton J. 1978. Possible teratogenicity of maternal fever. Lancet, 2, 634. 
Germain MA, Webster WS, Edwards MJ. 1985. Hyperthermia as a teratogen: parameters 
determining hyperthermia-induced head defects in the rat. Teratology, 31, 265-272. 
Graham JM, Edwards MJ, Edwards MJ. 1998. Teratogen update: gestational effects of 
maternal hyperthermia due to febrile illness and resultant patterns of defects in humans. 
Teratology, 58, 209-221. 

Hardy JD. 1961. Physiology of temperature regulation. Physiol Rev, 41, 521-606. 
Henderson J, Whittingham TA, Dunn T. 1997. A review of the acoustic output of modern 
diagnostic ultrasound equipment. BMUS Bull, 3, 10-14. 

Henderson J, Willson K, Jago JR, Whittingham TA. 1995. A survey of the acoustic outputs 
of diagnostic ultrasound equipment in current clinical use. Ultrasound Med Biol, 21, 699-705. 
Hendrickx AG, Stone GW, Hendrickson RV, Matayoshi K. 1979. Teratogenic effects of 
hyperthermia in the bonnet monkey (Macaca radiata). Teratology, 19, 177-182. 
Herman BA, Harris GR. 2002. Models and regulatory considerations for transient 
temperature rise during diagnostic ultrasound pulses. Ultrasound Med Biol, 28, 1217-1224. 
Horder MM, Barnett SB, Vella GJ, Edwards MJ, Wood AKW. 1998a. Ultrasound-induced 
temperature increase in the guinea pig fetal brain in utero: third-trimester gestation. 
Ultrasound Med Biol, 24, 1501-1510. 

65 



4 Ultrasound-induced heating and its biological consequences 

Horder MM, Barnett SB, Vella GJ, Edwards MJ, Wood AKW. 1998b. In vivo heating of the 
guinea-pig fetal brain by pulsed ultrasound and estimates of thermal index. Ultrasound 
Med Biol, 24, 1467-1474. 

Horder MM, Barnett SB, Edwards MJ. 1998c. Effects of pulsed ultrasound on sphenoid 
bone temperature and the heart rate in guinea-pig foetuses. Early Hum Dev, 52, 221-223. 
IEC. 2007. International Standard 60601-2-37 Medical Electrical Equipment - Part 2-37: 
Particular Requirements for the Basic Safety and Essential Performance of Ultrasonic 
Medical Diagnostic and Monitoring Equipment, Ed. 2.0. Geneva, Switzerland: Inter- 
national Electrotechnical Commission. 

IEC. 2010. International Standard 62359 Ultrasonics - Field Characterization - Test Methods 
for the Determination of Thermal and Mechanical Indices Related to Medical Diagnostic 
Ultrasonic Fields, Ed. 2.0. Geneva, Switzerland: International Electrotechnical Commission. 
Jago JR, Henderson J, Whittingham TA, Willson K. 1995. How reliable are manufacturers 
reported acoustic output data? Ultrasound Med Biol, 21, 135-136. 

Kimmel GL, Cuff JM, Kimmel CA, Heredia DJ, Tudor N, Silverman PM, et al. 1993. Skeletal 
development following heat exposure in the rat. Teratology, 47, 229-242. 
Kreshover SJ, Clough OW. 1953. Prenatal influences on tooth development. I. Artificially 
induced fever in rats. / Dent Res, 32, 565-577. 

Layde PM, Edmonds LD, Erickson JD. 1980. Maternal fever and neural tube defects. 
Teratology, 21, 105-108. 

Little BB, Ghali FE, Snell LM, Knoll KA, Johnston W, Gilstrap LC. 1991. Is hyperthermia 
teratogenic in humans? Am } Perinatol, 8, 185-189. 

Lubbers J, Hekkenberg RT, Bezemer RA. 2003. Time to threshold (TT), a safety parameter 
for heating by diagnostic ultrasound. Ultrasound Med Biol, 29, 755-764. 
Lyng H, Monge OR, Bohler PJ, Rofstad EK. 1991. Relationships between thermal dose and 
heat-induced tissue and vascular damage after thermoradiotherapy of locally advanced 
breast carcinoma. Int ] Hyperthermia, 7, 403^15. 

Macaulay JH, Randall NR, Bond K, Steer PJ. 1992. Continuous monitoring of fetal 
temperature by noninvasive probe and its relationship to maternal temperature, fetal 
heart rate, and cord arterial oxygen and pH. Obstet Gynecol, 79, 469^474. 
Mackowiak PA, Wasserman SS, Levine MM. 1992. A critical appraisal of 98.6 degrees 
F, the upper limit of the normal body temperature, and other legacies of Carl Reinhold 
August Wunderlich. / Am Med Assoc, 268, 1578-1580. 

Martin K. 2010. The acoustic safety of new ultrasound technologies. Ultrasound, 18, 1 10-1 18. 
Mazza S, Battaglia LF, Miller MW, Dewey WC, Edwards MJ, Abramowicz JS. 2004. The 
AT thermal dose concept 2: in vitro cellular effects. / Therm Biol, 19, 151-156. 
Mellette HC, Hurt BK, Askovitz SI, Horvath SM. 1951. Diurnal variations in body 
temperatures. / Appl Physiol, 3, 665-675. 

Miller MW, Ziskin MC. 1989. Biological consequences of hyperthermia. Ultrasound Med 
Biol, 15, 707-722. 

Miller MW, Nyborg WL, Dewey WC, Edwards MJ, Abramowicz JS, Brayman AA. 2002. 
Hyperthermic teratogenicity, thermal dose and diagnostic ultrasound during pregnancy: 
implications of new standards on tissue heating. Int } Hyperthermia, 18, 361-384. 
Moretti ME, Bar-Oz B, Fried S, Koren G. 2005. Maternal hyperthermia and the risk of 
neural tube defects in offspring. Systematic review and meta-analysis. Epidemiology, 16, 
216-219. 

66 



Ultrasound-induced heating and its biological consequences 4 



NA/NRC. 2006. Health Risks from Low Levels of Ionizing Radiation; BEIR VII Phase 2. 
Washington, DC: National Academy of Sciences/National Research Council. 
NCRP. 1990. Implementation of the Principle of as Low as Reasonably Achievable 
(ALARA) for Medical and Dental Personnel. Report no. 107. Bethesda, MD: National 
Council for Radiation Protection and Measurements. 

NCRP. 1992. Exposure Criteria for Medical Diagnostic Ultrasound: 1. Criteria Based on 
Thermal Mechanisms. Report no. 113. Bethesda, MD: National Council for Radiation 
Protection and Measurements. 

NCRP. 2002. Exposure Criteria for Medical Diagnostic Ultrasound: 1. Criteria Based on 
All Known Mechanisms. Report no. 140. Bethesda, MD: National Council for Radiation 
Protection and Measurements. 

Nelson TR, Fowlkes JB, Abramowicz JS, Church CC. 2009. Ultrasound biosafety 
considerations for the practicing sonographer/sonologist. / Ultrasound Med, 28, 139-150. 
Nielson NO. 1969. Teratogenic effects of hyperthermia. In Teratology, Bertelli A, Donati L 
(editors). Amsterdam, the Netherlands: Excerpta Medica Foundation. 
O'Brien WD, Deng CX, Harris GR, Herman BA, Merritt CR, Sanghvi N, et al. 2008. The risk 
of exposure to diagnostic ultrasound in postnatal subjects, thermal effects. / Ultrasound 
Med, 27, 517-535. 

Pennycuik PR. 1965. The effects of acute exposure to high temperatures on prenatal 
development in the mouse with particular reference to secondary vibrissae. Aust J Biol 
Sci, 18, 97-113. 

Pleet H, Graham JM, Smith DW. 1981 . Central nervous system and facial defects associated 
with hyperthermia at four to 14 weeks gestation. Paediatrics, 67, 785-789. 
Poswillo D, Nunnerly H, Sopher D, Keith J. 1974. Hyperthermia as a teratogenic agent. 
Ann R Coll Surg Engl, 55, 171-174. 

Raaphorst GP, Romano SL, Mitchell JB, Bedford JS, Dewey WC. 1979. Intrinsic differences 
in heat and/or X-ray sensitivity of seven mammalian cell lines cultured and treated under 
identical conditions. Cancer Res, 39, 396^01. 

Sapareto SA, Dewey WC. 1984. Thermal dose determination in cancer therapy, bit } Radiat 
Oncol, 10, 787-800. 

Sarge KD, Bray AE, Goodson ML. 1995. Altered stress response in testis. Nature, 
374, 126. 

Sasaki J, Yamaguchi A, Nabeshima Y, Shigemitsu S, Mesaki N, Kubo T. 1995. Exercise at 
high temperature causes maternal hyperthermia and fetal anomalies in rats. Teratology, 
51, 233-236. 

Shaw A, Pay NM, Preston RC. 1998. Assessment of the Likely Thermal Index Values for 
Pulsed Doppler Ultrasonic Equipment Stages 2 & 3: Experimental Assessment of Scanner/ 
Transducer Combinations. NPL Report CMAM 12. Teddington, UK: National Physical 
Laboratory. 

Shepard TH. 1982. Detection of human teratogenic agents. / Paediatr, 101, 810-815. 
Shepard TH. 1989. Catalogue of Teratogenic Agents. Baltimore, MD: John Hopkins 
University Press. 

Shiota K. 1982. Neural tube defects and maternal hyperthermia in early pregnancy: 
epidemiology in a human embryo population. Am ] Med Genet, 12, 281-288. 
Shiota K. 1988. Induction of neural tube defects and skeletal malformations in mice 
following brief hyperthermia in utero. Biol Neonate, 53, 86-97. 

67 



4 Ultrasound-induced heating and its biological consequences 



Skreb N, Frank A. 1963. Developmental abnormalities in the rat induced by heat-shock. 
/ Embryol Exp Morphol, 5, 311-323. 

Smith DW, Clarren SK, Harvey MAS. 1978. Hyperthermia as a possible teratogenic agent. 
/ Pediatr, 92, 878-883. 

Spraggett K, Fraser FC. 1982. Teratogenicity of maternal fever in women; a retrospective 
study. Teratology, 25, 75A. 

Suarez L, Felkner M, Hendricks K. 2004. The effect of fever, febrile illnesses, and heat 
exposures on the risk of neural tube defects in a Texas-Mexico border population. Birth 
Defects Res A Clin Mol Teratol, 70, 815-819. 

ter Haar G. 2010. The new British Medical Ultrasound Society Guidelines for the safe use 
of diagnostic ultrasound equipment. Ultrasound, 18, 50-51. 

Tikkanen J, Heinonen OP. 1991. Maternal hyperthermia during pregnancy and 
cardiovascular malformations in the offspring. Eur } Epidemiol, 7, 628-635. 
Upfold JB, Smith MSR, Edwards MJ. 1989. Quantitative study of the effects of maternal 
hyperthermia on cell death and proliferation in the guinea-pig brain on day 21 of 
pregnancy. Teratology, 39, 173-179. 

Walsh DA, Klein NW, Hightower LE, Edwards MJ. 1987. Heat shock and thermotolerance 
during early rat embryo development. Teratology, 36, 181-191. 

Webster WS, Edwards MJ. 1984. Hyperthermia and the induction of neural tube defects 
in mice. Teratology, 29, 417-425. 

WFUMB. 1992. World Federation for Ultrasound in Medicine and Biology Third 
Symposium in Medical Ultrasound. Issues and recommendations regarding thermal 
mechanisms for biological effects of ultrasound, Barnett SB, Kossoff G (editors). Ultrasound 
Med Biol, 18, 733-737. 

WFUMB. 1998. World Federation for Ultrasound in Medicine and Biology Symposium on 
Safety of Ultrasound in Medicine: conclusions and recommendations on thermal and non- 
thermal mechanisms for biological effects of ultrasound, Barnett SB (editor). Ultrasound 
Med Biol, 24(Special Issue)Sl, 1-55. 

Whittingham TA. 2000. The acoustic output of diagnostic machines. In The Safe Use of 
Ultrasound in Medical Diagnosis, ter Haar G, Duck FA (editors). 2nd Edition. London, 
UK: British Medical Ultrasound Society & The British Institute of Radiology, pp. 16-31. 
Ziskin MC. 2010. The thermal dose index. / Ultrasound Med, 29, 1475-1479. 



68 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 5 

Non-thermal effects of 
diagnostic ultrasound 

J. Brian Fowlkes 12 

1 Department of Radiology, University of Michigan, Ann Arbor, Ml, USA 
2 Department of Biomedical Engineering, University of Michigan, 

Ann Arbor, Ml, USA 



Summary 

• Mechanical effects, distinct from thermal effects, are those related to cavitation or 
other interactions of ultrasound with tissues not resulting in heating. 

• Some ultrasound imaging modes, such as acoustic radiation force impulse (ARFI) 
and shear wave imaging, rely on mechanical effects as a source of contrast in images. 

• Inertial cavitation can lead to high temperatures in and stresses around collapsing 
gas bubbles. 

• A variety of locations in tissues have demonstrated effects of ultrasound in animal studies 
but the incidence and significance of such effects in humans has not been determined. 

• Ultrasound contrast agents can provide significant diagnostic information and their 
safety profiles appear excellent based on clinical evidence although mechanisms for 
biological effects have been identified in animal studies. 

5.1 Introduction 

Diagnostic ultrasound has become one of the most widely used medical imaging 
modalities in the world. Its portability, relatively low cost and apparent lack of substantial 
biological effects have led to its use in a variety of medical disciplines. The benefits of 
ultrasound are well recognized, and its positive impact on healthcare is clear. Despite 
the large number of sonographic exams performed to date, there is no established casual 
relationship between clinical applications of diagnostic ultrasound and biologic effects on 
the patient or operator. 

As ultrasound imaging becomes ubiquitously available, it is important to understand 
how ultrasound propagation can affect tissue. Here we will concentrate on effects not 
directly related to temperature increases in tissue, which will be termed "non- thermal". 
While thermal effects can be readily understood as being associated with the absorption 

69 



Non-thermal 
effects arise 
from a number 
of mechanisms. 
Non-thermal 
effects may be 
cavitational or 
non-cavitational 
in origin 



Cavitation is the 
activation of 
small gas bodies 



5 Non-thermal effects of diagnostic ultrasound 



of ultrasound energy in tissue, non-thermal effects have a variety of source mechanisms, 
some of which will be discussed here. 



Gas bodies in 
tissue may occur 
naturally, or may 
be exogenous, 
as, for example, 
injected 
contrast agent 
microbubbles 



The non-thermal effects of ultrasound can be divided into two groups, cavitational and 
non-cavitational. Cavitation is the activation of small pockets of gas or vapour, commonly 
called gas bodies. In the case of diagnostic ultrasound, it is the variation of pressure in the 
ultrasound wave that activates these gas bodies. The source of gas bodies can be naturally 
occurring within the tissue or be exogenous, such as injectable microbubbles used as 
ultrasound contrast agents. Because of the significant emphasis placed on effects ascribed 
to cavitation, the activation of gas bodies by ultrasound, more attention will be devoted 
to this subject in this chapter. However, that does not diminish the potential significance 
of the other non-cavitational effects of ultrasound. 



Radiation 
force is related 
to the amount 
of energy 
absorbed by 
tissue 



5.2 Mechanical effects 

5.2.1 Radiation force 

The radiation force exerted on tissue is related to the amount of energy absorbed by the tissue. 

(Note that this absorbed energy may be converted to heat and is a source for thermal 
effects as well.) The time-average value of this force per unit volume of tissue is given by: 

2a I 

F v = — (5.1) 
c 

where a is the absorption coefficient of the medium, I is the acoustic intensity and c is the 
sound speed. The total force exerted on the tissue can then be given in terms of the total 
power absorbed from the ultrasound beam Was: 

W 



Ft = - 



(5.2) 



Note that for totally reflecting interfaces the radiation force is due to the momentum 
transfer required for the wave reflection. This condition doubles the local radiation force. 
It is the acoustic radiation force that is generally examined as a means of producing 
biological effects in tissue. 



Acoustic 
streaming 
results from 
radiation force 
in liquids 



Radiation 
force has 
a number 
of clinical 
applications, 
e.g. ARFI and 
SWEI 



Given sufficient power and tissue absorption (or reflection), there are appreciable forces 
that occur with the potential for biological effects. Another effect related to radiation force 
is acoustic streaming. Liquids can be caused to flow as a result of radiation forces. This 
effect has been used in diagnostic application for identification of fluid-filled cysts and 
their distinction from solid lesions (Nightingale et al, 1998). 

Overall, radiation force has seen a surge in medical application with the utilization of the 
wide dynamic range of the elastic modulus of tissues. Examples (with early references) 
of these methods include vibroacoustography (Fatemi and Greenleaf, 1998), shear wave 
elasticity imaging (SWEI) (Sarvazyan et al, 1998), ARFI imaging (Nightingale et al, 
2000) and supersonic shear imaging (Bercoff et al, 2004). A review of the topic provides 
a historic perspective and a more extensive collection of literature (Sarvazyan et al., 
2010). Given the actual use of radiation force in diagnostic procedures, the concept of 

70 



Non-thermal effects of diagnostic ultrasound 5 



bio-effects needs to be defined more in terms of deleterious effects, given that by 
definition ultrasound fields interact with the tissues. The effects can be temporary and 
without adverse consequence. 

Finally, the presence of bubbles can affect the local radiation force, increasing the effective 
absorption of the tissue and thereby increasing the radiation force. Radiation force will 
cause a bubble to move at speeds of -10 m s _1 in a cellular suspension and damage nearby 
cells by exposing them to high shear stresses near the bubble (Miller et al., 1991). 



5.2.2 Cavitation 

Cavitation, the interaction of ultrasound with gas bodies, has been investigated for over 
a century now, and although we have much understanding of the physical phenomenon, 
we have less knowledge about its inception in, and interactions with, tissue. The initiation 
of cavitation is affected by a number of parameters including acoustic field parameters 
[such as centre frequency, pulse repetition frequency (PRF) and pulse duration for pulsed 
ultrasound], tissue properties (such as density, viscosity and elasticity) and the size of 
any initial gas bodies (often referred to as the cavitation nuclei). The number and size 
of these nuclei may be a principal factor in the likelihood of biologic effects. Cavitation 
bubbles may be found only in small numbers and only at selected sites, and the applied 
acoustic field parameters, particularly the acoustic pressure, will control which nuclei 
can undergo cavitation. Bubble formation in animals has been modelled (Harvey et al., 
1944a,b; Yount, 1979) with some models applied to predict cavitation thresholds. 



Cavitation is often separated into types, with some confusion in the nomenclature (Church 
and Carstensen, 2001). However, inertial cavitation is one classification that is particularly 
useful and commonly used. Inertial cavitation occurs when the surrounding medium inertia 
controls the bubble motion (Flynn, 1975). When such cavitation occurs, the bubble collapse 
can be rapid, with large increases in the temperature inside and immediately surrounding 
the bubble, and causing significant mechanical stress to materials around the bubble. The 
prediction of conditions for inertial cavitation is the basis for the mechanical index (MI) 
(Holland and Apfel, 1989), to be discussed later, commonly used in the output display 
on diagnostic ultrasound systems. At very high acoustic pressures, beyond those used in 
diagnostic ultrasound, the bubbles can emulsify tissue (Parsons et al, 2006; Khokhlova 
et al, 2011). However, cavitation effects have also been observed with diagnostic pulses in 
fluids (Crum and Fowlkes, 1986; Holland et al., 1992; Carmichael et al, 1986). 



Inertial 
cavitation 
can lead 
to high 
temperature 
increases and 
significant 
stresses in 
tissue 



Cavitating bubbles can produce a variety of physical effects. As mentioned earlier, 
bubbles will affect the radiation force and will move in response to the field. Fluid flow 
immediately around the oscillating bubble, termed microstreaming, will subject cells to 
a high-velocity gradient as the fluid flow velocity will decrease away from the bubble. 
Shockwaves, produced by the bubble wall velocity exceeding the sound velocity in the 
gas and the tissue, will propagate through the tissue surrounding the bubble. Variations 
in tissue properties, and even within cells depending on the frequency components 
of the Shockwave, will result in differential motion and, consequently, local stress 
(Lokhandwalla et al., 2001; Lokhandwalla and Sturtevant, 2001). The temperature and 

71 



5 Non-thermal effects of diagnostic ultrasound 



pressure increases inside the bubble can produce free radicals, including the disassociation 
of water molecules from the water vapour in the collapsing bubble (Crum and Fowlkes, 
1986; Carmichael et al, 1986; Suslick and Flannigan, 2008; Flint and Suslick, 1991). Bubbles 
can collapse asymmetrically, resulting in microjets that have been observed in intravital 
microscopy in small vessels containing ultrasound contrast agent (Chen et al., 2011). 



5.3 Mechanical index 



Mechanical 
index estimates 
the potential 
for inertial 
cavitation 



The mechanical index, or MI, was adopted by the US Food and Drug Administration 
(FDA), American Institute of Ultrasound in Medicine and NEMA as a real-time output 
display to estimate the potential for inertial cavitation in vivo. The MI is given by: 

MI = p r ,/V^ ( 5. 3) 

where p , is the raref actional pressure (in MPa) of the acoustic field derated at 
0.3 dB (MHz cm) -1 and/ c is the centre frequency (in MHz) of the field. 



This index was based on a theoretical examination of the bubble collapse temperatures 
which could achieve 5000 K, a temperature at which free-radical generation can occur 
(Holland and Apfel, 1989). The MI is roughly proportional to the mechanical work that 
can be performed on a bubble in the rarefactional phase of the acoustic field. 



The MI is valid under conditions for the onset of inertial cavitation. Below an MI of -0.4, 
the physical conditions do not favour bubble growth even in the presence of a broad 
bubble nuclei distribution in the body, which is the assumption that is made for the MI 
formulation. 



5.4 Observations of bio-effects 



5.4.1 Bone 



Vascular 
damage has 
been seen near 
developing 
bone in small 
animal models 



The effect of ultrasound on bone growth has been studied for therapeutic applications. 
Although this is not a diagnostic application, it does use pulsed ultrasound to elicit 
the response. Pulsed ultrasound (PRF of 100-1000 Hz) has shown efficacy in fracture 
healing (Dyson and Brookes, 1983; Wang et al., 1994) and has accelerated the formation 
of the fracture callus in humans (Leung et al., 2004). The stimulation appears mediated 
by intracellular calcium signalling (Parvizi et al., 1999, 2002). It is not the megahertz 
carrier frequency commonly used in these experiments that is important since a 1 kHz 
squarewave signal produces similar chondrogenesis (Greenleaf et al., 2006). In addition, 
shock wave devices similar to lithotripters used to disrupt kidney stones accelerate bone 
growth and healing (Schaden, 1997; Wang et al., 2001). The pulsed nature of these devices 
makes it unlikely that heating is the mechanism for these effects. Vascular damage near 
developing bone has also been observed due to pulsed ultrasound (Dalecki et al., 1997c) 
and appeared only at a gestational age after bone formation began. The effect has been 
seen in developing mouse (Dalecki et al, 1999) and rat (Bigelow et al, 2007) foetuses. 
The requisite acoustic amplitude is above current output limits of diagnostic imaging 
devices. 



72 



Non-thermal effects of diagnostic ultrasound 5 



5.4.2 Lung 



There are several reports which indicate that ultrasound exposure, using diagnostically 
relevant acoustic parameters, can produce localized lung haemorrhage in animal models, 
such as that seen in Figure 5.1 ( Church et al, 2008). Results of these studies have previously 
been summarized (Church et al, 2008). The effects can be confounded by experimental 
methods where it has been shown that the acoustic impedance condition of the lung is 
important (O'Brien et al, 2002; Oelze et al, 2008). In fact, there is a suggestion for an 
alternative to the current MI for the case of the lung (Church and O'Brien, 2007). Table 5.1 
is a summary of experimental results involving lung eff arts in animal models (Church 
et al, 2008). Although the acoustic parameters used to produce these pulmonary eff ects are 
similar to those used in humans, the effects are focal and may not manifest themselves as 
a significant effect in human subjects. It could be argued that subjects with compromised 
lungs might be at greater risk, but there are no data to address this question. 



Lung 

haemorrhage 
in animal 
models has 
been observed 
as a result of 
ultrasound 
exposures 



5.4.3 Intestine 



Another location with naturally occurring gas bodies is the intestines. Although imaging 
through bowel gas is typically avoided, smaller gas bubbles contained within the intestine 
will not pose an imaging challenge but can still be activated by ultrasound. Using 
ultrasound near the current output limits as regulated by the FDA, effects have been seen 
in laboratory animals (Dalecki et al, 1995b; Miller and Gies, 1998, 2000). 



Effects have 
been seen in 
small animal 
models when 
gas-containing 
bowel is exposed 
to ultrasound 



5.4.4 Neurological development 

Marsal (2010) provides an overview of potential neurological development effects that 
have been examined specifically in epidemiological studies. Along with the review by 
Abramowicz et al. (2008b), the indication is that with the possible exception of an increase 




Figure 5.1. Appearance of subpleural haemorrhage (darker red area) in rat lung lobe 
following exposure to diagnostic ultrasound. Scale bar indicates 5.5mm. [Reproduced with 
permission from Church era/. (2008)]. 

73 



5 Non-thermal effects of diagnostic ultrasound 



Table 5.1. Summary of threshold data for lung haemorrhage. [Reproduced with permission 
from Church etal. (2008)]. 





Lung haemorrhage threshold results 


Nature of 
study 


Animal 


Frequency 
(MHz) 


Beamwidth 
(urn) 


PRF 
(kHz) 


Pulse 

duration 

(us) 


Exposure 
duration 
(s) 


p r , in 

situ 

(MPa) 


Threshold 
(Zachary et al, 
2001) 


Mouse 


2.8 


466 


1.0 


1.4 


10 


3.6 


Mouse 


5.6 


448 


1.0 


1.2 


10 


3.0 


Rat 


2.8 


466 


1.0 


1.4 


10 


2.3 


Rat 


5.6 


448 


1.0 


1.2 


10 


2.8 


Beam width 
(O'Brien et at., 
2001) 


Rat 


2.8 


470 


1.0 


1.1 


10 


3.6 


Rat 


2.8 


930 


1.0 


1.1 


10 


3.5 


Rat 


5.6 


310 


1.0 


1.1 


10 


3.5 


Rat 


5.6 


510 


1.0 


1.1 


10 


3.4 


Age dependence 
(O'Brien et al, 
2003) 


Pig,5d 


3.1 


610 


1.0 


1.2 


10 


3.6 


Pig, 39 d 


3.1 


610 


1.0 


1.2 


10 


5.8 


Pig, 58 d 


3.1 


610 


1.0 


1.2 


10 


2.9 


Threshold 
(O'Brien et al, 
2006) 


Rabbit 


5.6 


510 


1.0 


1.1 


10 


3.5 


Frequency (Child 

el wt., Lyyv ) 


Mouse 


3.7 


NR 


0.1 


1.0 


180 


1.4 


Threshold 
(Holland et al, 
1996) 


Rat 


4.0 


NR 


1.25 


1.0 


90 


2.0 


Rat 


4.0 


NR 


0.4 


1.0 


90 


2.5 


Pulse length 
(O'Brien et at., 
2003fl,b) 


Rat 


2.8 


470 


1.0 


1.3 


10 


3.1 


Rat 


2.8 


470 


1.0 


4.4 


10 


2.8 


Rat 


2.8 


470 


1.0 


8.2 


10 


2.3 


Rat 


2.8 


470 


1.0 


11.7 


10 


2.0 


jrrcULiciiL.y 
(Child et al, 
1990) 


Mouse 


1.1, U 


NR 


0.1 


10.0 


180 


0.4 


lVToi isp 


1.2 


NR 


0.1 


10.0 


180 


0.7 




2.3 U 


NR 


0.1 


10.0 


180 


0.6 


Mouse 


3.5, U 


NR 


0.1 


10.0 


180 


1.3 


Mouse 


3.7 


NR 


0.1 


10.0 


180 


1.0 


On time 
(Raeman et al, 
1993) 


Mouse 


1.2 


3500 


0.017 


10.0 


180 


1.1 


Threshold 
(Frizzell et al, 
1994) 


Mouse 


1.0 


1000 


0.1 


10.0 


180 


0.4 


Mouse 


JL.U 




JL.U 


1U.U 


9 A 


JL.O 


Exposure 
duration 
(Raeman et al, 
1996) 


Mouse 


2.3, U 


NR 


0.1 


10.0 


180 


0.7 


Mouse 


2.3, U 


NR 


0.1 


10.0 


20 


0.8 


Threshold 
(Baggs et al, 
1996) 


Pig 


2.3 


3000 


0.1 


10.0 


120 


0.9 


Threshold 
(Dalecki et al, 
1997a) 


Pig 


2.3 


3000 


0.1 


10.0 


120 


0.7 


Age dependence 
(Dalecki et al, 
1997b) 


Mouse, N 


1.15 


NR 


0.1 


10.0 


180 


0.6 


Mouse, J 


1.15 


NR 


0.1 


10.0 


180 


0.9 


Mouse, A 


1.15 


NR 


0.1 


10.0 


180 


0.7 



A indicates adult; J, juvenile; N, neonate; NR, not reported; and U, unfocused transducer. 



Non-thermal effects of diagnostic ultrasound 5 



incidence of non-right-handedness, there is insufficient evidence of an association. 
Marsal concluded that the epidemiological studies do not indicate abnormal neurological 
development as a consequence of ultrasound exposure. In the case of handedness, he 
pointed out some issues with the studies but indicated that they should not be disregarded. 

Continuing the topic, Marsal went on to discuss an animal study in which changes in 
neuronal migration were suggested (Ang et al, 2006). There has been some debate as to 
whether the experimental conditions can be related to those of human foetal exposure. While 
acoustic parameters used with the transvaginal probe selected for the study would be those 
of a commercial ultrasound system, the uninterrupted duration of the exposure, particularly 
with regard to the relative time for neuronal migration between species, and, importantly, the 
relative size of the brain to that of the ultrasound beam could lead to effects that might not 
manifest in clinical use. Nonetheless, the study points to the need for further investigation. 



5.4.5 Heart 



Application of ultrasound to the heart in an animal model has shown that radiation force 
can reduce the strength of contraction (Dalecki et al, 1993). Ultrasound pulses coincident 
with cardiac contraction of the frog heart did show this effect but only for a minimum pulse 
duration of 5 ms. This duration is orders of magnitude longer than typical diagnostic pulses 
with the possible exception of those now being used with acoustic radiation force to measure 
tissue elastic properties. However, no specific evaluation has been performed to investigate 
this potential effect with respect to imaging modes using acoustic radiation force. 



Evidence from 
small animals 
indicates that 
ultrasound 
can affect 
the strength 
of cardiac 
contractions 



5.4.6 Human perception 

Humans can perceive radiation force. For example, subjects were able to perceive 10 to 
100 ms pulses of 2 MHz ultrasound applied to the forearm when the power was greater than Humans 
20 W (Dalecki et al, 1995a). The foetus apparently will respond to the presence of ultrasound radiation 
during a diagnostic examination (Saeian et at, 1995). It is not clear if this is a response to the force 
sound produced at the PRF (Fatemi et al., 2001), which is in the audible range, or some other 
interaction with the foetus. It is expected that the sound pressure level is not substantial. 



5.4.7 Contrast 



There have been several reviews examining the safety of ultrasound contrast agents (ter 
Haar, 2009; Miller et al, 2008a). Microbubble contrast agents provide an obvious source 
of cavitation nuclei that would not normally be found in tissue. Based on animal studies, 
ultrasound exposure of tissues containing microbubble contrast agents can result in 
bio-effects, including haemolysis (Dalecki et al, 1997d), capillary rupture (Miller and 
Quddus, 2000; Wible et al, 2002), endothelial damage (Kobayashi et al, 2002, 2003), 
cardiac arrhythmias (van der Wouw, 2000; Li et al, 2003, 2004; Dalecki et al, 2005) and 
effects in the kidney (Miller et al, 2007a,b, 2008b, 2009, 20Wa,b; Williams et al, 2007) and 
the pancreas (Miller et al, 2011). In a review of this research (Miller et al, 2008a), the 
conclusion drawn was that "the use of high MI values (>0.8) involves rapid gas-body 
destruction with a potential for bio-effects (e.g., PVCs), whereas bio-effects have not been 
observed at low values of (p/ -Jf <0.2), which involve minimal gas-body destruction". 

75 



The safety 
and efficacy 
of ultrasound 
microbubble 
contrast 
agents is under 
continuous 
review 



5 Non-thermal effects of diagnostic ultrasound 



Certainly, this is consistent with the common practice of low-MI imaging where the 
output is maintained below levels expected to produce significant contrast disruption as 
this would eliminate the signal needed for imaging. The exception would be the use of 
contrast replenishment where the agent is purposely disrupted to monitor its return to the 
tissue. There are specific uses for such imaging, and it will be important to consider the 
benefit of the procedure against any potential risk. Recently, there has been a concerted 
effort to summarize and add to the evidence on the safety of ultrasound contrast agents 
in echocardiology. The FDA issued a "black box" warning for contrast agent use. This 
precipitated a considerable response from the scientific community investigating ultrasound 
contrast safety. Main et al. (Main, 2009; Main et al, 2009) summarized the results of these 
studies and concluded that there is substantial evidence to support the safe and effective 
use of ultrasound contrast for the current indicated clinical uses in echocardiology. 

5.5 Conclusions 

Given the considerations indicated here and the literature on the topic of non-thermal 
effects of ultrasound similar to that used in diagnostic ultrasound, effects have been 
observed in animal models, although the significance of these and the relationship to the 
clinical application of ultrasound is unclear. Certainly, there is not sufficient epidemiologic 
evidence to conclude a causal relationship between diagnostic ultrasound and adverse 
bio-effects in patients. This is the same conclusion drawn in other reviews (Abramowicz 
et al., 2008a,b). However, there are reasons why we may actually need to increase acoustic 
output to achieve diagnostic information that is of value to the patient. For example, 
acoustic radiation force used to evaluate tissue elasticity may require outputs exceeding 
those currently regulated by the FDA to be able to perform the procedure in deeper tissues. 
It is important that we consider this objectively from the standpoint of the benefit to the 
patient versus the risk, if any, that exists. We should be addressing these questions out of 
an obligation to do what is correct for the patient and should even accept some additional 
risk if necessary. It can be argued that the bigger risk to the patient may reside in the 
decision to obtain the information by some other imaging modality or to not perform any 
imaging procedure. This could subject the patient to much greater risks. 

References 

Abramowicz JS, Barnett SB, Duck FA, Edmonds PD, Hynynen K, Ziskin MC. 2008a. Fetal 
thermal effects of diagnostic ultrasound. / Ultrasound Med, 27, 541-559. 
Abramowicz JS, Fowlkes JB, Skelly AC, Stratmeyer ME, Ziskin MC. 2008b. Conclusions 
regarding epidemiology for obstetric ultrasound. / Ultrasound Med, 27, 637-644. 
Ang Jr ES, Gluncic V, Duque A, Schafer ME, Rakic P. 2006. Prenatal exposure to ultrasound 
waves impacts neuronal migration in mice. Proc Natl Acad Sci USA, 103, 12903-12910. 
Baggs R, Penney DP, Cox C, Child SZ, Raeman CH, Dalecki D, et al. 1996. Thresholds for 
ultrasonically induced lung hemorrhage in neonatal swine. Ultrasound Med Biol, 22, 119-128. 
Bercoff J, Tanter M, Fink M. 2004. Supersonic shear imaging: a new technique for soft 
tissue elasticity mapping. IEEE Trans Ultrason Ferroelectr Freq Control, 51, 396^09. 
Bigelow TA, Miller RJ, Blue JPJ. 2007. Hemorrhage near fetal rat bone exposed to pulsed 
ultrasound. Ultrasound Med Biol, 33, 311-317. 

76 



Non-thermal effects of diagnostic ultrasound 5 



Carmichael AJ, Mossoba MM, Riesz P, Christman CL. 1986. Free radical production in 
aqueous solutions exposed to simulated ultrasonic diagnostic conditions. IEEE Trans 
Ultrason Ferroelectr Freq Control, UFFC 33, 148-155. 

Chen H, Brayman AA, Kreider W, Bailey MR, Matula TJ. 2011. Observations of translation 
and jetting of ultrasound-activated microbubbles in mesenteric microvessels. Ultrasound 
Med Biol, 37, 2139-2148. 

Child SZ, Hartman CL, Schery LA, Carstensen EL. 1990. Lung damage from exposure to 
pulsed ultrasound. Ultrasound Med Biol, 16, 817-825. 

Church CC, Carstensen EL. 2001. "Stable" inertial cavitation. Ultrasound Med Biol, 27, 1435-1437. 
Church CC, Carstensen EL, Nyborg WL, Carson PL, Frizzell LA, Bailey MR. 2008. The 
risk of exposure to diagnostic ultrasound in postnatal subjects - nonthermal mechanisms. 
/ Ultrasound Med, 27, 565-592. 

Church CC, O'Brien Jr WD. 2007. Evaluation of the threshold for lung hemorrhage by 
diagnostic ultrasound and a proposed new safety index. Ultrasound Med Biol, 33, 810-818. 
Crum LA, Fowlkes JB. 1986. Acoustic cavitation generated by microsecond pulses of 
ultrasound. Nature, 319, 52-54. 

Dalecki D, Child SZ, Raeman CH, Carstensen EL. 1995a. Tactile perception of ultrasound. 
JAcoust SocAm, 97, 3165-3170. 

Dalecki D, Child SZ, Raeman CH, Cox C. 1999. Hemorrhage in murine fetuses exposed to 
pulsed ultrasound. Ultrasound Med Biol, 25, 1139-1144. 

Dalecki D, Child SZ, Raeman CH, Cox C, Carstensen EL. 1997a. Ultrasonically induced 
lung hemorrhage in young swine. Ultrasound Med Biol, 23, 777-781. 
Dalecki D, Child SZ, Raeman CH, Cox C, Penney DP, Carstensen EL. 1997b. Age dependence 
of ultrasonically induced lung hemorrhage in mice. Ultrasound Med Biol, 23, 767-776. 
Dalecki D, Child SZ, Raeman CH, Penney DP, Mayer R, Cox C, et al. 1997c. Thresholds for 
fetal hemorrhages produced by a piezoelectric lithotripter. Ultrasound Med Biol, 23, 287-297. 
Dalecki D, Keller BB, Raeman CH, Carstensen EL. 1993. Effects of ultrasound on the frog 
heart. I: thresholds for changes in cardiac rhythm and aortic pressure. Ultrasound Med 
Biol, 19, 385-390. 

Dalecki D, Raeman CH, Child SZ, Carstensen EL. 1995b. Intestinal hemorrhage from 
exposure to pulsed ultrasound. Ultrasound Med Biol, 21, 1067-1072. 

Dalecki D, Raeman CH, Child SZ, Cox C, Francis CW, Meltzer RS, et al. 1997d. Hemolysis 
in vivo from exposure to pulsed ultrasound. Ultrasound Med Biol, 23, 307-313. 
Dalecki D, Rota C, Raeman CH, Child SZ. 2005. Premature cardiac contractions produced 
by ultrasound and microbubble contrast agents in mice. Acoust Res Lett Online, 6, 221-226. 
Dyson M, Brookes M. 1983. Stimulation of bone repair by ultrasound. Ultrasound Med 
Biol, 9(Suppl. 2), 61-66. 

Fatemi M, Greenleaf JF. 1998. Ultrasound-stimulated vibro-acoustic spectrography. 
Science, 280, 82-85. 

Fatemi M, Ogburn PL, Greenleaf JF. 2001. Fetal stimulation by pulsed diagnostic 
ultrasound. / Ultrasound Med, 20, 883-889. 

Flint EB, Suslick KS. 1991. The temperature of cavitation. Science, 253, 1397-1399. 

Flynn HG. 1975. Cavitation dynamics. I. A mathematical formulation. / Acoust Soc Am, 57, 

1379-1396. 

Frizzell LA, Chen E, Lee C. 1994. Effects of pulsed ultrasound on the mouse neonate: hind 
limb paralysis and lung hemorrhage. Ultrasound Med Biol, 20, 53-63. 

77 



5 Non-thermal effects of diagnostic ultrasound 



Greenleaf JF, Argadine HM, Bolander ME. 2006. 1 kHz vibration stimulates ATDC5 
chondrocytes. In Fifth International Symposium on Therapeutic Ultrasound, Clement GT, 
McDannold NJ, Hynynen K (editors). New York, NY: American Institute of Physics, pp. 49-53. 
Harvey EN, Barnes KK, McElroy WD, Whitely AH, Pease DC, Cooper KWJ. 1944a. Bubble 
formation in animals. I. Physical factors. / Cell Comp Physiol, 24, 1-22. 
Harvey E, Barnes K, McElroy W, Whitely A, Pease D, Cooper K. 1944b. Bubble formation in 
animals. II. Gas nuclei and their distribution in blood and tissues. / Cell Comp Physiol, 24, 23-34. 
Holland CK, Apfel RE. 1989. An improved theory for the prediction of microcavitation 
due to pulsed ultrasound. IEEE Trans Ultrason Ferroelectr Freq Control, 36, 204-208. 
Holland CK, Deng CX, Apfel RE, Alderman JL, Fernandez LA, Taylor KJ. 1996. Direct 
evidence of cavitation in vivo from diagnostic ultrasound. Ultrasound Med Biol, 22, 917-925. 
Holland CK, Roy RA, Apfel RE, Crum LA. 1992. In vitro detection of cavitation induced 
by a diagnostic ultrasound system. IEEE Trans Ultrason Ferroelectr Freq Control, 39, 95-101. 
Khokhlova TD, Canney MS, Khokhlova VA, Sapozhnikov OA, Crum LA, Bailey MR. 
2011. Controlled tissue emulsification produced by high intensity focused ultrasound 
shock waves and millisecond boiling. JAcoust SocAm, 130, 3498-3510. 
Kobayashi N, Yasu T, Yamada S, Kudo N, Kuroki M, Kawakami M, et al. 2002. Endothelial 
cell injury in venule and capillary induced by contrast ultrasonography. Ultrasound Med 
Biol, 28, 949-956. 

Kobayashi N, Yasu T, Yamada S, Kudo N, Kuroki M, Miyatake K, et al. 2003. Influence of 
contrast ultrasonography with perflutren lipid microspheres on microvessel injury. Circ 
J, 67, 630-636. 

Leung K-S, Lee W-S, Tsui H-F, Liu P-L, Cheung W-H. 2004. Complex tibial fracture outcomes 
following treatment with low-intensity pulsed ultrasound. Ultrasound Med Biol, 30, 389-395. 
Li P, Armstrong WF, Miller DL. 2004. Impact of myocardial contrast echocardiography 
on vascular permeability: comparison of three different contrast agents. Ultrasound Med 
Biol, 30, 83-91. 

Li P, Cao LQ, Dou CY, Armstrong WF, Miller DL. 2003. Impact of myocardial contrast 
echocardiography on vascular permeability: an in vivo dose response study of delivery 
mode, pressure amplitude and contrast dose. Ultrasound Med Biol, 29, 1341-1349. 
Lokhandwalla M, McAteer JA, Williams Jr JC, Sturtevant B. 2001. Mechanical haemolysis in 
shock wave lithotripsy (SWL): II. In vitro cell lysis due to shear. Phys Med Biol, 46, 1245-1264. 
Lokhandwalla M, Sturtevant B. 2001. Mechanical haemolysis in shock wave lithotripsy 
(SWL): I. Analysis of cell deformation due to SWL flow-fields. Phys Med Biol, 46, 413-437. 
Main ML. 2009. Ultrasound contrast agent safety: from anecdote to evidence. ]ACC 
Cardiovasc Imaging, 2, 1057-1059. 

Main ML, Goldman JH, Grayburn PA. 2009. Ultrasound contrast agents: balancing safety 
versus efficacy. Expert Opin Drug Saf, 8, 49-56. 

Marsal K. 2010. Exposure to ultrasound in utero: epidemiology and relevance of neuronal 
migration studies. Ultrasound Med Biol, 36, 1221-1223. 

Miller DL, Averkiou MA, Brayman AA, Everbach EC, Holland CK, Wible Jr JH, et al. 2008a. 
Bioeffects considerations for diagnostic ultrasound contrast agents. / Ultrasound Med, 27, 611-632. 
Miller DL, Dou C, Sorenson D, Liu M. 2011. Histological observation of islet hemorrhage 
induced by diagnostic ultrasound with contrast agent in rat pancreas. PLoS One, 6, e21617. 
Miller DL, Dou C, Wiggins RC. 2008b. Frequency dependence of kidney injury induced 
by contrast-aided diagnostic ultrasound in rats. Ultrasound Med Biol, 34, 1678-1687. 

78 



Non-thermal effects of diagnostic ultrasound 5 



Miller DL, Dou C, Wiggins RC. 2007a. Doppler mode pulse sequences mitigate glomerular 
capillary hemorrhage in contrast-aided diagnostic ultrasound of rat kidney. IEEE Trans 
Ultrason Ferroelectr Freq Control, 54, 1802-1810. 

Miller DL, Dou C, Wiggins RC, Wharram BL, Goyal M, Williams AR. 2007b. An in vivo 
rat model simulating imaging of human kidney by diagnostic ultrasound with gas-body 
contrast agent. Ultrasound Med Biol, 33, 129-135. 

Miller DL, Dou C, Wiggins RC. 2009. Glomerular capillary hemorrhage induced in rats by 
diagnostic ultrasound with gas-body contrast agent produces intratubular obstruction. 

Ultrasound Med Biol, 35, 869-877. 

Miller DL, Dou CY, Wiggins RC. 2010a. Contrast-enhanced diagnostic ultrasound causes 

renal tissue damage in a porcine model. / Ultrasound Med, 29, 1391-1401. 

Miller DL, Dou CY, Wiggins RC. 2010b. Gas body efficacy for glomerular capillary 

hemorrhage induced by diagnostic ultrasound in rats. IEEE Trans Biomed Eng, 57, 167-174. 

Miller DL, Gies RA. 1998. Gas-body-based contrast agent enhances vascular bioeffects of 

1.09 MHz ultrasound on mouse intestine. Ultrasound Med Biol, 24, 1201-1208. 

Miller DL, Gies RA. 2000. The influence of ultrasound frequency and gas-body composition 

on the contrast agent-mediated enhancement of vascular bioeffects in mouse intestine. 

Ultrasound Med Biol, 26, 307-313. 

Miller DL, Quddus J. 2000. Diagnostic ultrasound activation of contrast agent gas bodies 
induces capillary rupture in mice. Proc Natl Acad Sci USA, 97, 10179-10184. 
Miller DL, Thomas RM, Williams AR. 1991. Mechanisms for hemolysis by ultrasonic 
cavitation in the rotating exposure system. Ultrasound Med Biol, 17, 171-178. 
Nightingale KR, Kornguth PJ, Trahey GE. 1998. The use of acoustic streaming in breast 
lesion diagnosis: a clinical study. Ultrasound Med Biol, 25, 75-87. 

Nightingale KR, Nightingale RW, Palmeri ML, Trahey GE. 2000. A finite element model 
of remote palpation of breast lesions using radiation force: factors affecting tissue 
displacement. Ultrason Imaging, 22, 35-54. 

O'Brien WD Jr, Kramer JM, Waldrop TG, Frizzell LA, Miller RJ, Blue JP, et al. 2002. 
Ultrasound-induced lung hemorrhage: role of acoustic boundary conditions at the 
pleural surface. ] Acoust SocAm, 111, 1102-1109. 

O'Brien WDJ, Simpson DG, Frizzell LA, Zachary JF. 2001. Superthreshold behavior 
and threshold estimation of ultrasound-induced lung hemorrhage in adult rats: role of 
beamwidth. IEEE Trans Ultrason Ferroelectr Freq Control, 48, 1695-1705. 
O'Brien WDJ, Simpson DG, Frizzell LA, Zachary JF. 2003a. Threshold estimates and 
superthreshold behavior of ultrasound-induced lung hemorrhage in adult rats: role of 
pulse duration. Ultrasound Med Biol, 29, 1625-1634. 

O'Brien WDJ, Simpson DG, Ho MH, Miller RJ, Frizzell LA, Zachary JF. 2003b. 
Superthreshold behavior and threshold estimation of ultrasound-induced lung 
hemorrhage in pigs: role of age dependency. IEEE Trans Ultrason Ferroelectr Freq Control, 
50, 153-169. 

O'Brien Jr WD, Yang Y, Simpson DG, Frizzell LA, Miller RJ, Blue Jr JP, et al. 2006. Threshold 
estimation of ultrasound-induced lung hemorrhage in adult rabbits, and comparison of 
thresholds in rabbits, rats and mice. Ultrasound Med Biol, 32, 1793-1804. 
Oelze ML, Miller RJ, Blue Jr JP, Zachary JF, O'Brien Jr WD. 2008. Estimation of the acoustic 
impedance of lung versus level of inflation for different species and ages of animals. 
J Acoust SocAm, 124, 2340-2352. 

79 



5 Non-thermal effects of diagnostic ultrasound 



Parsons JE, Cain CA, Abrams GD, Fowlkes JB. 2006. Pulsed cavitational ultrasound 
therapy for controlled tissue homogenization. Ultrasound Med Biol, 32, 115-129. 
Parvizi J, Parpura JF, Greenleaf JF, Bolander ME. 2002. Calcium signaling is necessary for 
ultrasound-stimulated aggrecan synthesis by rat chondrocytes. / Orthop Res, 20, 51-57. 
Parvizi J, Wu CC, Lewallen DG, Greenleaf JF, Bolander ME. 1999. Low- in tensity 
ultrasound stimulates proteoglycan synthesis in rat chondrocytes by increasing aggrecan 
gene expression. / Orthop Res, 17, 488-494. 

Raeman CH, Child SZ, Carstensen EL. 1993. Timing of exposures in ultrasonic hemorrhage 
of murine lung. Ultrasound Med Biol, 19, 507-512. 

Raeman CH, Child SZ, Dalecki D, Cox C, Carstensen EL. 1996. Exposure-time dependence of the 
threshold for ultrasonically induced murine lung hemorrhage. Ultrasound Med Biol, 22, 139-141. 
Saeian K, Weintraub R, Hagen- Ansert S, Sahn C, Shiota T, Kenny A, et al . 1 995 . Increased fetal 
activity and heart rate during and immediately after pulsed Doppler echocardiography. 
Echocardiography, 12, 71-77. 

Sarvazyan AP, Rudenko OV, Nyborg WL. 2010. Biomedical applications of radiation 
force of ultrasound: historical roots and physical basis. Ultrasound Med Biol, 36, 1379-1394. 
Sarvazyan AP, Rudenko OV, Swanson SD, Fowlkes JB, Emelianov SY. 1998. Shear wave 
elasticity imaging: a new ultrasonic technology of medical diagnostics. Ultrasound Med 
Biol, 24, 1419-1435. 

Schaden W. 1997. Clinical experience with shock wave therapy of pseudarthrosis, delayed 
fracture healing, and cement-free endoprothesis loosening. In Extracorporeal Shock 
Waves in Orthopaedics, Siebert W, Buch M (editors). Heidelberg, Germany: Springer- 
Verlag, pp. 137-148. 

Suslick KS, Flannigan DJ. 2008. Inside a collapsing bubble: sonoluminescence and the 
conditions during cavitation. Annu Rev Rhys Chem, 59, 659-683. 

ter Haar G. 2009. Safety and bio-effects of ultrasound contrast agents. Med Biol Eng 
Comput, 47, 893-900. 

van der Wouw PA. 2000. Premature ventricular contractions during triggered imaging 
with ultrasound contrast. J Am Soc Echocardiogr, 13, 288-294. 

Wang C, Chen H, Yang K. 2001. Treatment of nonunions of long bone fractures with shock 
waves. Clin Orthop Relat Res, 387, 95-101. 

Wang SJ, Lewallen DG, Bolander ME, Chao EYS, Illstrup DM. 1994. Low-intensity ultrasound 
treatment increases strength in a rat femoral fracture model. / Orthop Res, 12, 40^7. 
Wible JH, Galen KP, Wojdyla JK, Hughes MS, Klibanov AL, Brandenburger GH. 2002. 
Microbubbles induce renal hemorrhage when exposed to diagnostic ultrasound in 
anesthetized rats. Ultrasound Med Biol, 28, 1535-1546. 

Williams AR, Wiggins RC, Wharram BL, Goyal M, Dou C, Johnson KJ, et al. 2007. Nephron 
injury induced by diagnostic ultrasound imaging at high mechanical index with gas body 
contrast agent. Ultrasound Med Biol, 33, 1336-1344. 

Yount DE. 1979. Skins of varying permeability: a stabilization mechanism for gas cavitation 
nuclei. ] Acoust Soc Am, 65, 1429-1439. 

Zachary JF, Frizzell LA, Norrell KS, Blue JP, Miller RJ, O'Brien WDJ. 2001. Temporal and 
spatial evaluation of lesion reparative responses following superthreshold exposure of rat 
lung to pulsed ultrasound. Ultrasound Med Biol, 27, 829-839. 



80 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 6 

Radiation force and its possible 

biological effects 

Hazel C. Starritt 

Royal United Hospital, Bath, UK 



Summary 

• Radiation force is generated within a material in an acoustic field. 

• Bio-effects attributed to radiation forces have been observed experimentally in tissue; 
these include physical effects and sensory effects. 

• The volume force generated in diagnostic fields causes biological fluids to stream. 

• Any radiation force effects produced at diagnostic exposure levels are transitory. 

• Only under extreme conditions has permanent damage to cells been observed. 

• Accelerated tissue repair resulting from ultrasound exposure has been attributed to 
radiation force. 

6.1 Introduction 

Traditionally the mechanisms producing ultrasound bio-effects have been classed as 
either thermal or cavitational. However, there are reports in the literature of bio-effects 
occurring in situations where heating and cavitation are unlikely to be contributing 
factors. Radiation force is the most probable mechanism occurring in these circumstances, 
since low-level radiation forces are exerted whenever an ultrasound beam passes through 
tissue. Useful reviews can be found in the literature (Duck, 1998; Barnett, 1998; Sarvazyan 
et ol., 2010). This chapter examines aspects of radiation force effects in the context of 
ultrasound safety. Although the forces resulting from ultrasound propagation are very 
small indeed, nevertheless some of the effects that they produce can be easily observed 
under appropriate conditions. One example is acoustic streaming and another is the force 
exerted on the target of a radiation force balance. 



81 



6 Radiation force and its possible biological effects 



6.2 Radiation force 



An ultrasound 
beam generates 
a force at an 
interface 



6.2.1 Radiation force at a boundary 

Before considering the radiation forces experienced throughout a three-dimensional 
volume of fluid or tissue, a simpler situation will be discussed: that of an ultrasound beam 
passing through a fluid and reaching a solid surface. The surface could, for example, 
be a target in a radiation force balance, or in the body it could be an interface between 
soft tissue and bone, or tissue and air in the lungs. A force is generated at the surface. 
This tends to push it away from the source of ultrasound. The strength of this push will 
depend on the details of the surface and of the beam. The shape of the surface, its size 
in comparison with the acoustic beam, the angle of incidence and the extent to which it 
absorbs or reflects the ultrasound beam will influence the magnitude and direction of 
the force. In addition, the intensity or power in the ultrasound beam is a determining 
factor. So long as the surface extends outside the ultrasound beam, the total radiation 
force exerted on the surface is proportional to the total acoustic power in the beam. 
Measurement of this force is a standard method of determining acoustic power. 



Spatial variations 
in the radiation 
pressure at a 
surface result in 
shear forces in 
an absorber 



The force is the integral of the radiation pressure across the surface. At each point the 
radiation pressure depends on the ratio of the intensity in the beam to the velocity of 
sound in the fluid. For a totally absorbing surface perpendicular to the beam, the local 
radiation pressure is explicitly the local acoustic intensity divided by the speed of sound. 
Hence the radiation pressure will be greatest on the axis of the beam where the intensity 
is greatest, and will decrease towards the edges of the beam as the intensity falls off. The 
radiation pressure profile does, in fact, vary across the beam in the same manner as the 
intensity profile. As a result of this variation in radiation pressure at the surface, shear 
stress will be generated within an absorber. 



A volumetric force 
is established by 

energy absorption 
in an ultrasound 
field 



6.2.2 Radiation forces in a volume 

The next stage is to consider the more complicated situation of what happens within a 
volume of tissue or fluid when an ultrasound beam passes through it. An ultrasound 
beam passing through a material such as soft tissue will lose energy to the material by 
a process of absorption. As a result, a volumetric internal force that acts in the direction 
of wave propagation is generated in the material. For a plane wave, the force per unit 
volume is given by Iodic, where c is the speed of sound, I is the local intensity and a is 
the attenuation coefficient. It varies throughout the field as the intensity varies. So, for 
an unscanned beam (Doppler and M-mode) the volume force is greater at the focus and 
on the axis of the beam than elsewhere. In addition, because of its dependence on the 
absorption coefficient, the radiation force is greater at higher acoustic frequencies, and 
varies because of tissue inhomogeneity and non-linear enhancement (see Chapter 2). The 
situation is very complex, and this makes it difficult to carry out a complete theoretical 
analysis of the forces that are likely to be generated. 



The universal presence of radiation force in all ultrasonic fields has led to its exploitation 
in new methods for elastography. For example, acoustic radiation force impulse imaging 

82 



Radiation force and its possible biological effects 6 



uses short-duration (typically less than 1ms) acoustic radiation forces to generate 
localized displacements in tissue, which are detected and used as the basis for the image 
(Nightingale et al., 2001, 2002). Shear wave elasticity imaging characterizes the shear 
modulus of the medium by mapping the velocity of shear waves generated by a short- 
duration acoustic radiation force (Bercoff et al., 2002). These and other similar methods 
all derive information on the elastic properties of tissue from the detection of small 
displacements caused by radiation force, whilst still operating at low enough spatial 
average intensities to lie within international safety constraints for diagnostic ultrasound. 



Radiation force 
elastography 
exploits the 
universal presence 
of radiation force 
in all ultrasound 
fields 



Radiation force is experienced only during the passage of an acoustic pulse. Between 
pulses no force is generated. For pulsed Doppler and pulse-echo applications the 
magnitude of the force depends on the pulse-average intensity and not on the time- 
average intensity. For continuous wave ultrasound systems such as physiotherapy units 
and foetal heart monitors, the force is proportional to the time-average intensity. 



Radiation force 
is experienced 
only during the 
passage of an 
acoustic pulse 



6.3 Acoustic streaming 



6.3.1 Introduction 

Acoustic streaming results from the generation of a force field in a liquid in the 
direction of wave propagation. It is a directly observable, bulk movement of liquid 
away from a transducer that occurs as a result of the absorption of acoustic energy from 
an ultrasound beam. All absorption processes contribute to this effect, including shear 
and bulk viscosity, relaxation and excess absorption due to non-linear propagation. 
There have been a number of attempts to predict streaming in plane, progressive waves 
(Eckart, 1948; Nyborg, 1998) under bounded and free-field conditions although none 
have modelled the situation in diagnostic ultrasound fully (Kamakura et al., 1995; 
Mitome et al., 1995). Nevertheless, generalized expressions derived from theoretical 
models give a good indication of the way in which the velocity of a stream depends 
upon the characteristics of the acoustic beam and the propagation fluid. In particular, 
the flow velocity is proportional to the acoustic intensity, the radius of the acoustic 
beam and the amplitude attenuation coefficient associated with all processes of acoustic 
loss. The maximum velocity reached in a fluid is limited by viscous forces and by the 
geometric boundaries of the fluid space. 



Acoustic 
streaming can 
be generated 
in fluids by an 
ultrasound beam 



6.3.2 Investigation of acoustic streaming in vitro 

Acoustic streaming can easily be demonstrated using most diagnostic ultrasound scanners. 
Measurements of streaming induced using commercial equipment reported by 
Starritt et al. (1989) yielded a maximum streaming velocity in water of 14cms~ in the fi 
eld of a diagnostic scanner operated in pulsed Doppler mode. Figure 6.1 shows a 
stream produced in water using a single element 3.5 MHz diagnostic transducer operating 
in pulsed Doppler mode. In imaging fields the streaming velocities reported were 
lower, of the order of \ cm s~ . Streaming has also been observed in vitro in human blood 
and in human serum albumin solution. Zauhar et al. (2006) compared acoustic streaming 
in amniotic fluid with that in water. They demonstrated that the speed was greater in 

83 



Streaming occurs 
in diagnostic 
ultrasound beams 



6 Radiation force and its possible biological effects 




Figure 6.1. Picture of a stream produced in water in the field of a 3.5 MHz transducer operating 
in pulsed Doppler mode. (Reproduced from Starritt etal., 1991, with permission). 

amniotic fluid than in water, because of the difference in absorption coefficient. However, 
when using high-amplitude ultrasound pulses, the speeds were very similar, the force 
being dominated by non-linear absorption effects (see Chapter 2). 



Streaming can 
be generated 
at powers as 
low as 1mW 



More sensitive means, such as magnetic resonance imaging, have been used to investigate 
streaming at low acoustic powers and within restricted spaces (Starritt et ah, 2000). In water, 
streaming velocities as low as 0.1 mm s -1 were measured, generated by acoustic powers 
as low as 1 mW. This is at the lower end of the power levels available from commercial 
ultrasound equipment operating in Doppler mode (Duck and Henderson, 1998; Martin, 
2010). Streaming was also detected within the pores of a coarse sponge. These results 
imply that streaming is probably occurring within fluid spaces in vivo during diagnostic 
ultrasound examinations more frequently than may be appreciated in practice. It is not 
normally detected, because of the low velocities involved. 



Streaming has 
been observed 
in vivo 



Although not often reported in the literature, streaming in fluid spaces in vivo may be 
easily observed during scanning when operating at higher intensities. Fluid movement 
has been reported in breast cysts by Nightingale et al. (1995) and proposed as a diagnostic 
tool for distinguishing solid from fluid-filled cysts. 



It is very 
unlikely indeed 
that acoustic 
streaming occurs 
in cells and 
extracellular 
spaces 



It has been postulated that fluids may be caused to stream within the extracellular spaces 
and even within cells themselves. There is no experimental evidence to suggest that this 
happens, and it seems to be extremely unlikely to occur. The creation of a stream requires 
a finite fluid volume, and it is strongly inhibited by the cohesive forces from the volume 
boundary, and the viscous forces of the fluid. The physical environment of cellular fluids 
is such that these cohesive forces will be sufficient to prevent fluid movement caused by 
radiation force. 



84 



Radiation force and its possible biological effects 6 



6.3.3 Non-linear propagation 



The propagation of high-amplitude pulses, such as those used for diagnostic ultrasound, 
is accompanied by losses that result in extra absorption of ultrasound energy (see Chapter 
2) . This in turn increases the eff active att enuation coefficient, altering the volumetric force 
and any fluid streaming arising from it. If tissue lies behind a fluid path, radiation forces 
on this tissue can be enhanced. This arises because a shocked wave, carrying harmonic 
components generated during propagation through a fluid, impinges on tissue and 
results in enhancement of the absorption. Similarly enhanced forces may also operate on 
tissue in the absence of such a fluid path, but such effects are much weaker. 



Non-linearity can 
lead to enhanced 
absorption 
resulting in 
increased 
radiation force 
and streaming 



One factor in any consideration of the significance of acoustic streaming in clinical 
situations is the time period over which a stream develops. Starritt et al. (1991) found that 
in water the shortest rise times to establish the stream were of the order of a few hundred 
milliseconds for unscanned pulsed beams and up to a few seconds for imaging beams. 
Hartley (1997) reported rise times in water and blood of 200 ms and 80 ms, respectively. 
It is clear that acoustic streaming becomes established very quickly when an ultrasound 
beam is applied to a fluid. 



Streaming 
can become 
established in 
under a second 



6.4 Radiation force in fluids and tissues 

Streaming is just one potential outcome of the generation of internal stresses in a material 
caused by the passage of an ultrasound beam. The forces causing fluids to stream will also 
be present in soft tissue, but tissue is not free to move in the same way as a fluid. Some 
of the effects on tissues and fluids that may arise from these forces are discussed below. 



The forces causing 
streaming in fluids 
are also present 
within tissue 



6.4.1 Effects of radiation force on fluids and cell suspensions 



In all bulk fluids and cell suspensions, the fluid will move as a result of acoustic streaming. 
However, it is unlikely that the process of gentle stirring, for example of amniotic fluid 
or urine in vivo, will present a biological hazard. Shear forces will occur at the boundary 
of a stream, but these are unlikely to cause damage to cells in suspension. A simple 
calculation shows that for a streaming velocity of 10 cms -1 the shear is about 10 Pa at 
the boundary of a 2 mm stream, which is well below the threshold of 150 kPa for lysis of 
erythrocytes. 



Shear forces 
arising from 
streaming are 
unlikely to 
damage cells 



Acoustic streaming has been suggested as the mechanism responsible for observed 
changes in diffu sion across a planar lipid bilayer membrane (Pohl et al., 1993). The thickness 
of the unstirred layer near the membrane was reduced in the presence of ultrasound, 
particularly on the side nearest to the transducer. This is a significant finding because 
unstirred boundary layers have an essential role to play in transport across biological 
membranes. The phenomenon of phonophoresis, that is, the enhancement of diffusion 
rates caused by ultrasound, is best explained by the effect that ultrasound has on the 
boundary fluid layers. These effects may be further enhanced in the presence of bubbles. 
In addition, Pohl et al. (1995), suggest that acoustic streaming may explain an observed 
effect of ultrasound on the ability of red blood cells to form aggregates. 



Streaming can 
alter the thickness 
of unstirred 
boundary layers 



85 



6 Radiation force and its possible biological effects 



6.4.2 Effects of radiation force on soft tissue 

A number of papers report observations of radiation force effects on soft tissues. These 
may be grouped into two categories, as either physical effects or sensory effects. 



Radiation 
force has been 
associated with 
some physical 
effects in tissue 



6.4.2.1 Physical effects on tissue 

Several papers in the literature report physical effects on tissue which the authors 
believe may be better explained as radiation force effects than thermal or cavitation 
effects. Lizzi et al. (1981) have reported blanching of the choroid of the eye prior to the 
onset of thermal damage. It has been suggested that this occurred due to radiation force 
causing compression of the blood vessels. Dalecki et al. (1997a) used an experimental 
lithotripter to deliver ultrasound pulses to the abdomen of pregnant mice. Pulse 
amplitudes were in the diagnostic range, but the pulse powers used were higher. The 
foetal tissue showed evidence of haemorrhage, but only where the soft tissue was near 
to developing bone or cartilage. The authors suggest that this could result from the 
relative motion between ossified bone and surrounding soft tissue, caused by radiation 
force on the bone. 



The accelerated 
tissue repair 
observed in an 
acoustic field is 
probably due to 
mechanical forces 



There have been a number of reports of accelerated healing of bone fractures in vivo using 
low intensity pulsed ultrasound (Kristiansen et al, 1997; Heckman et al., 1994). Although 
the precise biophysical mechanism is unknown, it has been suggested that it arises from 
the application of mechanical force to the cellular system. An associated alteration in gene 
expression has been reported (Yang et al, 1996; Parvisi et ah, 1999; see also Chapter 7). 
Enhancement of soft-tissue regeneration has also been reported using low intensity 
therapeutic ultrasound (Dyson et al., 1968, 1970). The effect, which is greatest during 
the early stages of regeneration, was not attributed to heating due to the low intensities 
employed. 



Radiation 
forces can alter 
neurosensory 
responses 



6.4.2.2 Sensory effects on tissue 

Several papers have suggested radiation force as the biophysical mechanism for 
neurosensory responses. Dalecki et al. (1995) have demonstrated that it is possible to feel 
the radiation forces that are exerted on the skin by an ultrasound beam. Also reported by 
Dalecki et al. (1997b) was a decrease in aortic pressure caused by ultrasound insonation of 
frog hearts. The authors demonstrated that radiation force was responsible by showing 
an equivalent effect when the beam was incident on a total absorber in contact with the 
surface of the heart. A number of papers have reported that the auditory nerve may be 
directly stimulated by ultrasound (Magee and Davies, 1993). The mechanism is unknown 
but we can speculate that it is the direct effect of the varying force field across the neural 
structures. 



Preliminary 
observations of 
altered neuronal 
migration have 
been attributed to 
radiation force 



6.4.2.3 Developmental effects 

Ang et al. (2006) reported evidence that exposure to ultrasound from a clinical scanner 
caused partial inhibition of neural migration in the embryonic cerebral cortex of mice. In 
the absence of evidence suggesting heating or cavitation as the cause, radiation force was 
proposed as being the most likely mechanism. 

86 



Radiation force and its possible biological effects 6 



6.5 Pulsed radiation force as a bio-effects mechanism 



Radiation force is experienced only during the time a pulse is passing through tissue 
and there is no force between pulses. The magnitude of the force in modes such as 
pulse echo and pulsed Doppler depends on the pulse-average intensity rather than 
the time-average intensity. In continuous wave applications such as physiotherapy or 
foetal heart monitoring the tissue experiences a steady force dependent on the time- 
averaged intensity. 



In pulsed 
ultrasound fields 
the radiation 
force depends on 
the pulse-average 
intensity 



Traditionally, biological effects that show a dependence on time-average intensity are 
interpreted as being thermal in origin and those that depend on pulse-average intensity or 
pulse amplitude are explained in terms of cavitation. Once we start to consider radiation 
force as a possible mechanism it becomes more difficult to separate the dependence on 
exposure factors. Although radiation force effects are experienced only during a pulse, 
and are therefore dependent on pulse amplitude or pulse-average intensity, some of the 
outcomes, like streaming and some of the sensory effects, rely on an integration of the 
force over time. In this respect they resemble thermal effects. However, the time scales can 
be very different. Acoustic streaming for example is established almost instantaneously, 
in time scales of less than a second, whereas tissue temperature takes tens of seconds to 
increase. 



Radiation force 
effects occur over 
a shorter time 
scale than thermal 
effects 



Table 6.1 shows the minimum pulse lengths and number of pulses required in order to 
produce the radiation force effects described in the papers reviewed above. It shows that 
some effects can be produced and observed following only a single pulse of ultrasound, 
while others require the force to be repeated over a number of pulses. 

From this table it can be seen that, for example, 200 pulses of about 10 ps duration were 
required to produce haemorrhage in foetal mouse tissue. Choroid blanching, however, 
occurred with a single pulse about 100 ps in duration and the cardiac response in frogs 
was also seen with a single long pulse, 5 ms in duration. In order to sense ultrasound on 
the skin a repetitive stress is required and similarly, whilst local fluid movement must be 
induced by a single pulse, there needs to be a repeated effect before it manifests itself as 
bulk streaming. 

Table 6.1. Minimum ultrasonic pulse duration and number of pulses for radiation force- 
induced effects. 



Effect 



Pulse length Number of pulses 



Haemorrhage at bone/soft tissue 
interface in mice 

Choroid blanching 



= 10 us 
100 us 



200 
1 



Some radiation 
force effects are 
manifest in a 
single acoustic 
pulse 



Tactile sensation 
Cardiac response in fro£ 
Fluid movement 



1 ms 
5 ms 
0.5 us 



Repetitive 
1 



87 



6 Radiation force and its possible biological effects 



6.6 Cellular mechano-transduction 



Cellular mechano- 
transduction 
pathways will 
respond to 
radiation force in 
ways that are yet 
to be understood 



Cells sense and respond to external forces, by a wide range of mechanisms, yet to be fully 
understood (Huang et ah, 2003). They do so in order to protect themselves from shear, 
probably the most threatening force, and to adapt their function to altered mechanical 
environments. It is believed that the cell membrane can sense mechanical forces by 
means of the molecular agents integrins. This mechano-sensing links extracellular forces 
to the cytoskeleton, and thus can initiate a cellular response both by gene expression 
and biochemical change. A broadly homogeneous force may be amplified as a result of 
local cell-matrix or cell-cell adhesions, with amplification factors of 100 being suggested. 
The effect of haemodynamic shear on the vascular endothelium is the mechanical effect 
subject to the most detailed study (Van Bavel, 2007; Davies, 1995). The critical level of 
fluid shear stress for a variety of biological responses is about lPa. Force thresholds 
associated with other experiments suggest a response threshold of about InN. It may 
be expected, therefore, that cells can detect, and may respond to, radiation force during 
many ultrasound diagnostic procedures. 



6.7 Conclusion 

Radiation force effects provide a possible explanation for ultrasound bio-effects which 
appear to be non-thermal and non-cavitational in nature. In adult tissue these forces 
are highly unlikely to be significant compared with the tensile strength of tissue, even 
that of weak adult tissue. However, in the embryonic stage tissue does not have the 
structural strength that develops in later foetal and adult life since the intercellular 
matrix has yet to develop. The period of organogenesis, between 3.5 weeks and 8 weeks 
gestation in humans, is a period during which cell differentiation and migration is 
occurring, and it is possible that the developing foetus may be more vulnerable to 
mechanical stress at this time. However, there is insufficient evidence to know whether 
or not the passage of an ultrasound beam could exert sufficient radiation force to 
cause permanent displacement of cells. The effect of these known radiation forces on 
cellular mechano-transduction is likewise very unclear. It is therefore important to 
keep the potential for bio-effects arising from radiation forces in mind, particularly 
when ultrasound scanning is carried out during the first trimester. As previously 
recommended, it is prudent to reduce the exposure whenever this can be done without 
compromising diagnostic information. 



References 

Ang E, Gluncic V, Duque A, Schafer M, Rakic P. 2006. Prenatal exposure to ultrasound 
waves impacts neuronal migration in mice. PNAS, 103, 12903-12910. 
Barnett S (editor). 1998. Other non-thermal mechanisms: acoustic radiation force and 
streaming. In Conclusions and Recommendations on Thermal and Non-thermal 
Mechanisms for Biological Effects of Ultrasound. Proceedings of the World Federation for 
Ultrasound in Medicine and Biology Symposium on Safety of Ultrasound in Medicine. 
Ultrasound Med Biol, 24(Suppl. 1), S23-S28. 



88 



Radiation force and its possible biological effects 6 



Bercoff J, Chaffai S, Tanter M, Fink M. 2002. Ultraf ast imaging of beamformed shear waves 
induced by the acoustic radiation force in soft tissue: application to transient elastography. 
In Proc 2002 IEEE Ultrasonics Symposium, Yuhas DE, Schneider SC (editors). New York, 
NY: IEEE, pp. 1899-1902. 

Daleki D, Raeman CR, Child SZ, Carstensen EL. 1997a. Effects of pulsed ultrasound on 
the frog heart: III. The radiation force mechanism. Ultrasound Med Biol, 23, 275-285. 
Daleki D, Child SZ, Raeman CR, Penney DP, Meyer R, Cox C, et al. 1997b. Thresholds 
for fetal haemorrhages produced by a piezoelectric lithotripter. Ultrasound Med Biol, 23, 
287-297. 

Daleki D, Child SZ, Raeman CH, Carstensen EL. 1995. Tactile perception of ultrasound. 
JAcoust SocAm, 97, 3165-3170. 

Davies PF. 1995. Flow-mediated endothelial mechanotransduction. Physiol Rev, 75, 
519-556. 

Duck FA, 1998. Acoustic streaming and radiation pressure in diagnostic applications. 
In Safety of Diagnostic Ultrasound, Barnett SB, Kossoff G (editors). Carnforth, UK: 
Parthenon, pp. 87-98. 

Duck FA, Henderson J. 1998. Acoustic output of modern ultrasound equipment: is it 
increasing? In Safety of Diagnostic Ultrasound, Barnett SB, Kossoff G (editors). Carnforth, 
UK: Parthenon, pp. 15-25. 

Dyson M, Pond JB, Joseph J, Warwick R. 1968. The stimulation of tissue regeneration by 
means of ultrasound. Clin Sci, 35, 273-285. 

Dyson M, Pond JB, Joseph J, Warwick R. 1970. Stimulation of tissue regeneration by 
pulsed plane-wave ultrasound. IEEE Trans Sonics Ultrason, SU 17, 133-139. 
Eckart C. 1948. Vortices and streams caused by sound waves. Physiol Rev, 73, 68-76. 
Hartley CJ. 1997. Characteristics of acoustic streaming created and measured by pulsed 
Doppler ultrasound. IEEE Trans Ultrason Ferroelectr Freq Control, 44, 1278-1285. 
Heckman JD, Ryaby JP, McCabe J, Frey J, Kilcoyne RF. 1994. Acceleration of tibial fracture- 
healing by non-invasive low-intensity pulsed ultrasound. / Bone Joint Surg, 76, 6-43. 
Huang H, Kamm RD, Lee RT. 2003. Cell mechanics and mechanotransduction: pathways, 
probes and physiology. Am ] Cell Physiol, 287, Cl-Cll. 

Kamakura T, Matsuda K, Kumamoto Y. 1995. Acoustic streaming induced in focused 
Gaussian beams. JAcoust Soc Am, 97, 2740-2746. 

Kristiansen TK, Ryaby JP, Frey JJ, Roe LR. 1997. Accelerated healing of distal radial 
fractures with the use of specific, low-intensity ultrasound. A multicenter, prospective, 
randomized, double-blind, placebo-controlled study. / Bone Joint Surg Am, 79, 961-973. 
Lizzi FL, Coleman DJ, Driller J, Franzen LA, Leopold M. 1981. Effects of pulsed ultrasound 
on ocular tissue. Ultrasound Med Biol, 7, 245-252. 

Magee TR, Davies AH. 1993. Auditory phenomena during transcranial Doppler insonation 
of the basilar artery. / Ultrasound Med, 12, 747-750. 

Martin K. 2010. The acoustic safety of new ultrasound technologies. Ultrasound, 18, 110-118. 
Mitome H, Kozuka T, Tuziuti T. 1995. Effects of nonlinearity in development of acoustic 
streaming. Jpn J Appl Phys, 34, 2584-2589. 

Nightingale KR, Kornguth PJ, Walker WF, McDermott BA, Trahey GE. 1995. A novel 
ultrasonic technique for differentiating cysts from solid lesions: preliminary results in the 
breast. Ultrasound Med Biol, 21, 745-751. 



89 



6 Radiation force and its possible biological effects 



Nightingale KR, Palmeri ML, Nightingale R, Trahey G. 2001. On the feasibility of remote 
palpation using acoustic radiation force. J Acoust SocAm, 110, 625-634. 
Nightingale KR, Soo MS, Nightingale R, Trahey G. 2002. Acoustic radiation force impulse 
imaging: in vivo demonstration of clinical feasibility. Ultrasound Med Biol, 28, 227-235. 
Nyborg N. 1998. Acoustic streaming. In Non-linear Acoustics, Hamilton MF, Blackstock 
DT (editors). New York, NY: Academic Press, pp. 207-231. 

Parvisi JP, Wu CC, Lewallen DG, Greenleaf JF, Bolander ME. 1999. Low- in tensity 
ultrasound stimulates proteoglycan synthesis in rat chondrocytes by increasing aggrecan 
gene expression. / Orthop Res, 17, 488-494. 

Pohl P, Antonenko YN, Rosenfeld E. 1993. Effect of ultrasound on the pH profiles in 
the unstirred layers near planar bilayer lipid membranes measured by microelectrodes. 
Biochim Biophys Acta, 1152, 155-160. 

Pohl EE, Rosenfeld EH, Pohl P, Millner R. 1995. Effects of ultrasound on agglutination 
and aggregation of human erythrocytes in vitro. Ultrasound Med Biol, 21, 711-719. 
Sarvazyan AP, Rudenko OV, Nyborg WL. 2010. Biomedical applications of radiation 
force of ultrasound: historical roots and physical basis. Ultrasound Med Biol, 36, 1379-1394. 
Starritt HC, Duck FA, Humphrey VF. 1989. An experimental investigation of streaming in 
pulsed diagnostic ultrasound beams. Ultrasound Med Biol, 15, 363-373. 
Starritt HC, Duck FA, Humphrey VF. 1991. Forces acting in the direction of propagation 
in pulsed ultrasound fields. Phys Med Biol, 36, 1465-1474. 

Starritt HC, Hoad CL, Duck FA, Nassiri DK, Summers IR, Vennart W. 2000. Measurement 
of acoustic streaming using magnetic resonance. Ultrasound Med Biol. 26, 321-333. 
Van Bavel E. 2007. Effects of shear stress on endothelial cells: possible relevance for 
ultrasound applications. Prog Biophys Mol Biol, 93, 374-383. 

Yang KH, Parvizi J, Wang SJ, Lewallen DG, Kinnick RR, Greenleaf JF, et al. 1996. Exposure 
to low intensity ultrasound increases the aggrecan gene expression in a rat femur fracture 
model. / Orthop Res, 14, 802-809. 

Zauhar G, Duck FA, Starritt HC. 2006. Comparison of the acoustic streaming in amniotic 
fluid and water in medical ultrasonic beams. Ultraschall Med/Eur } Ultrasound, 14, 152-158. 



90 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 7 

Bio-effects — cells and tissues 

Gail ter Haar 

Institute of Cancer Research, Sutton, UK 



Summary 

Laboratory experiments carried out in vivo and in vitro allow the following conclusions 
to be drawn: 

• In the absence of acoustic cavitation, no cell lysis or loss of reproductive integrity has 
been seen. 

• No effects have been seen when embryos or foetuses are exposed to ultrasound 
in utero if the temperature is kept below 41.5 °C. 

• Observed reduction in birth weight following ultrasound exposure is probably due 
to changes in maternal physiology. 

• Mechanisms which lead to effects observed in vitro may not occur in the same way 
in vivo. 

• Unless an ultrasonic beam is scaled appropriately, a significantly larger fraction of an 
experimental animal is exposed than would be the case for a human foetus. 

7.1 Introduction 

The interaction of an ultrasonic beam with the tissues through which it passes is an 
essential prerequisite for production of a diagnostic ultrasound scan. It is, after all, the 
scattered beam that is used to form an image. It has long been known that ultrasound can 
induce change in biological tissues, and this is the basis for physiotherapy applications 
where beneficial changes (which may often be reversible) in cellular function are sought, 
and in surgery, where cell killing is required. The requirement for the safe use diagnostic 
ultrasound is that any cellular changes that may occur are reversible and do not 
constitute a hazard to the individual being scanned. In the quest for information about 
safety, a number of different experimental models and ultrasonic exposures have been 
investigated. The resulting literature is somewhat confusing, but a number of conclusions 
may be drawn, as outlined in this chapter. For a fuller treatment of this topic, the reader is 
directed to fuller reviews such as those published by NCRP (2002), Miller (2007), ter Haar 
(2007) or AGNIR (2010). 



91 



7 Bio-effects — cells and tissues 



7.2 Studies of isolated cells 



The relative 
importance of 
physical damage 
mechanisms is 
different when 
cells are held in 
vitro or are in 
intact tissues in 
vivo 



Experiments conducted with single cells can yield useful information about ultrasonically 
induced changes produced under closely defined exposure conditions, in well controlled 
physiological environments. A number of endpoints have beenused. These include "gross" 
effects such as lysis, loss of reproductive ability and damage to cellular ultrastructure, 
and more subtle effects such as altered growth patterns, chromosomal and functional 
changes. Care must be taken, however, in extrapolating from results seen with cells 
exposed to ultrasound in vitro to effects that might be expected when cells are subjected 
to ultrasound in vivo. The aqueous nutrient media required for these cell experiments 
have a lower threshold for acoustic cavitation induction, and a lower acoustic absorption 
coefficient than would be expected in vivo. Cavitation is therefore more likely, arid 
thermal damage less likely, to occur under these circumstances. The presence of liquid in 
close contact with the cells may also enhance any effects due to acoustic streaming. Thus, 
the relative importance of physical mechanisms that may result in damage are different 
when cells are held in vitro from when they exist in intact tissues in vivo. Provided this is 
realized, useful results may be obtained from such models. 



Cavitation can 
lead to cell lysis 



7.2.1 Cell lysis 

There is no doubt that ultrasound exposures can lyse cells. Cavitation has been shown 
by a number of people to be the major mechanism in producing complete cellular 
destruction (see, for example, Kaufman et ah, 1977; Morton et ah, 1982; Hallow et ah, 
2006; Lai et ah, 2007). It is not clear that ultrasound can produce cell lysis in the absence 
of cavitation. It has been shown that the proportion of cells lysed depends on the 
cellular concentration, high cellular concentrations exhibiting a lower proportion of 
cells disrupted than low ones (see for example Elwart et ah, 1988; Brayman et ah, 1996a). 
This cell density effect may be due in part to the higher respiratory consumption of 
dissolved oxygen, with its concomitant release of CO, into the suspension medium, 
thus reducing the probability of cavitation effects (Brayman et ah, 1992; Carstensen 
et ah, 1993). Lysis appears to be an immediate consequence of ultrasonic exposure, 
rather than a delayed effect. Cells actively undergoing division (in mitosis) are more 
susceptible to being lysed by a given ultrasonic exposure than those at other stages in 
the cell cycle (Clarke and Hill, 1969). 



Cells subjected 
to simultaneous 
heat and 
ultrasound 
exposure may 
lose their ability 
to divide 



7.2.2 Loss of reproductive integrity 

The clonogenic assay is a commonly-used measure of toxicity in radiobiology. It is used 
to assess the ability of a cell to divide and produce viable progeny following an insult. 
In general, it appears that cells which are not lysed by an ultrasound exposure go on 
to divide in the same way as their untreated counterparts (Bleaney et ah, 1972; Morton 
et ah, 1982). The exception to this appears to lie with cells that are heated (41-45 °C) while 
being exposed to ultrasound (Li et ah, 1977; ter Haar et ah, 1980; Feril and Kondo, 2004). 
These exhibit a loss in reproductive ability that is greater than if the cells were subjected to 
heat alone. The mechanism for this is not fully understood, but is thought to be non- 
thermal and non-cavitational in origin (Morton et ah, 1983). 

92 



Bio-effects — cells and tissues 7 



7.2.3 Ultrastructural changes 

There is a wealth of literature concerned with the effects of ultrasound on cellular 
ultrastructure. A variety of changes have been reported, many of them sublethal or 
reversible. 



The extracellular membrane is an obvious target for study. Ultrasonically induced 
effects have usually manifested themselves as changes in permeability to ion transport. 
There are several examples of this, including the sublethal alteration in the thymocyte 
plasma membrane that leads to a decrease in potassium concentration following 1 W cm -2 
irradiation in vitro at 1.8 MHz (Chapman, 1974) and the reversible increase in calcium 
uptake in fibroblasts shown by Mortimer and Dyson (1988) [1MHz, 0.5-1 .0W cm" 2 1 ]. 
These effects have been under extensive investigation more recently with an upsurge of 
interest in ultrasound enhancement of drug and gene delivery (ter Haar, 2007; Deckers 
and Moonen, 2010). 



Ultrasound 
can affect 
extracellular 
membrane 
permeability 



Organelle damage following ultrasound exposure has been demonstrated using electron 
microscopy. Mitochondria are the most frequently affected. Lysosomal damage has 
been demonstrated following exposure to intact tissues, with the consequent release 
of lysosomal enzymes. However, it is not clear whether this is a direct or indirect 
eff act of the ultrasound on the lysosomes (Dvorak and Hrazdira, 1966; Hrazdira, 
1970; Taylor and Pond, 1972). Dilated rough endoplasmic reticulum (RER) and some 
irregular intracellular lesions in addition to membrane and mitochondrial effects 
have been observed in exposures at intensities lying above the cavitation threshold 
(Harvey et ah, 1975). 



Ultrasound 
affects 
mitochondria 
and lysosomes, 
but not the cell 
nucleus 



In general, in the absence of cavitation, the cell nucleus appears to be unaffected by 
ultrasonic exposures. The only lesions that have been reported are slit-like vacuoles at the 
nuclear membrane (ter Haar et al., 1979). 



7.2.4 DNA and chromosomal effects 



Ultrasound can degrade DNA in solution. It appears that cavitation is necessary to 
achieve this, and that the damage is due to the hydrodynamic shear stresses, free-radical 
formation or excessive heating that may accompany cavitation events (Thacker, 1973; 
Miller and Thomas, 1995, 1996). As discussed in Chapter 5, such cavitation activity is 
unlikely to occur during diagnostic ultrasound exposures. 



DNA damage 
can only be 
induced by 
ultrasound in 
the presence 
of acoustic 
cavitation 



Considerable effort has been expended in looking for ultrasonically induced 
chromosomal aberrations and for sister chromatid exchange (SCE). The evidence is 
strong that even up to quite high intensities (100 W cm 2 ) ultrasound does not produce 
chromosomal damage (see Rott, 1981; EFSUMB, 1994 for comprehensive reviews). 
Although there have been occasional reports that ultrasound may induce chromosomal 
effects, these have never been independently substantiated by a second investigation, 
and the vast majority of the literature has yielded negative results. There may, however, 
be some synergistic effect when an ultrasound exposure follows X-irradiation, but 
not when it precedes it (810kHz, 3 Wcm 2 ; Kunze-Muhl, 1981). 

93 



Ultrasound 
does not 
produce 
chromosome 
damage 



7 Bio-effects — cells and tissues 



There are SCE analysis is a frequently used assay for the effect of potentially mutagenic agents 
conflicting Q n mammalian cells (Latt and Schreck, 1980; Gebhart, 1981). Liebeskind et al. (1979) 
^CE increase reported that diagnostic ultrasound might be capable of inducing SCEs in vitro. 

This stimulated a considerable amount of work, with the majority of investigators 
reporting negative findings even for intensities up to 3.0 W cirr 2 (3.15 MHz cw). The 
significance of an increase in SCEs for the cell or whole organism is not properly 
understood. 



7.2.5 Functional changes 



Ultrasound 
enhances 
collagen 
synthesis 



Ultrasound exposures may act either to stimulate or inhibit cellular function. For example, 
ultrasound irradiation of human fibroblasts in vitro may act to increase protein synthesis. 
Ultrasound has been shown to have considerable effect on the synthesis of collagen by 
fibroblasts both in vitro and in vivo (Dyson and Smalley, 1983; Webster et al., 1979). Harvey 
et al. (1975) have demonstrated that when primary diploid human fibroblasts were 
irradiated with 3 MHz ultrasound at an intensity of 0.5 W cirr 2 in vitro, the amount of 
protein synthesized was increased. Electron microscopic examination of irradiated cells 
revealed that, in comparison with control cells, there were more free ribosomes, more 
dilatation of RER, more cytoplasmic vacuolation, more autophagic vacuoles and more 
damage to lysosomal membranes and mitochondria. Subsequent work from the same 
group (Webster et al., 1978, 1980) has shown that cavitation may be involved in producing 
this stimulation of collagen synthesis. It has also been shown that ultrasound can stimulate 
the release of histamine by mast cell degranulation (Fyfe and Chahl, 1982), possibly by an 
increase of calcium ion transport across their membranes (Mortimer and Dyson, 1988). 



Ultrasound 
affects 
cellular 
movement 
and mobility 



Other observed functional changes have been associated with cellular motion. In time lapse 
photomicrography studies, ultrasonically induced changes in cellular movement that 
lasted for several generations were reported (Liebeskind et al., 1982). The electrophoretic 
mobility of cells may be affected by ultrasound (Taylor and Newman, 1972) although 
this reflects a change in surface cell density arising from an increase in cellular volume. 
This appears only to occur in vitro in association with cavitation events (Mummery, 1978; 
Joshi et al, 1973). 



7.3 Studies of multicellular organisms 

A wide range of tissue models has been used for the study of the effects of ultrasound on 
multicellular structures. These include plants, insects, fish, small and large animals and 
man. Only mammalian studies will be highlighted here, although useful information has 
also been derived from the other models. 



7.3.1 Bone effects 

The principle cause for concern when bone lies in an ultrasonic beam lies with 
periosteal heating since attenuation of energy by bone is too high to allow significant 



94 



Bio-effects — cells and tissues 7 



penetration at diagnostic frequencies. This heating is likely to provide the limiting 
factor in physiotherapy or hyperthermia treatments as the periosteum is rich in nerve 
endings. There is also the possibility of significant bone heating with pulsed Doppler 
examinations at the maximum available output levels. In the aware human, with normal 
pain sensitivity, excessive periosteal heating is likely to lead to pain. If the treatment is 
stopped when pain is felt, then it is probable that damage will be avoided. Of potential 
pulsed Doppler examinations, those in obstetrics give the most grounds for concern on 
these thermal grounds as proliferating tissue has been shown to be especially susceptible 
to thermal injury. In laboratory animals there have been measurements of biological 
significant temperature rises (>2°C) close to the skull bone as a result of ultrasonic 
exposures (Carstensen et al., 1990; Duggan et al., 1995; Bosward et al., 1993; Horder et al., 
1998; see Chapter 4). 



Bone may be 
heated by 
ultrasound 



There are some reports that ultrasound can accelerate bone healing (Claes and 
Willie, 2007, Chang et al, 2002, Duarte, 1983). The mechanism for this is not clear. 
Experimental studies of fracture in rat fibulae indicate that ultrasonic irradiation during 
the inflammatory and early proliferative phases of repair accelerates and enhances 
healing. Direct ossification, with little cartilage production, is seen. Treatment in the late 
proliferative phase, however, was found to be disadvantageous, cartilage growth being 
stimulated, with delay to bony union (Dyson and Brookes, 1983). In their study, it was 
found that 1.5 MHz was more effective than 3.0 MHz (I , 0.5 W cm 2 , 2ms:8ms, 5min), 

satp ' 'I' 

and so a non-thermal effect is suggested. This finding has been repeated by a number 
of authors. Pilla et al. (1990) showed that the strength of that of intact bone was reached 
in ultrasonically treated rabbit fibulae 17 days after osteotomy as compared to 28 days 
for control animals (1.5 MHz, I ata 0.03 Worr 2 , 200 ps:800 ps, 20min daily). Heckman et al. 
(1994) demonstrated similar acceleration of healing in a human clinical trial. They treated 
open fractures of the tibial shaft, and found a significant reduction in the time needed 
to achieve clinical and radiographic healing (96 ±4.9 days for the ultrasonically treated 
group, 154±13.7 days for the control group) (1.5 MHz, I ata 0.03 Wcnr 2 , 200ps:800ps, 
daily, starting within 7 days of fracture). There appears to be evidence that it is not only 
the time at which treatment is started that is important, but also the dose level. Too high 
an intensity can lead to inhibition of protein synthesis, or, at worst to deleterious effects. 
Tsai et al. (1992) found that whereas 0.5 W cm -2 (I t ) significantly accelerated bone repair, 
1.0cm -2 (I ) suppressed the repair process (1.5 MHz, 200 ps, 5-20min daily). Reher et al. 
(1997) found in an in vitro study of the effect of ultrasound exposure on mouse calveria 
bone, that whereas 0.1 W cm -2 (3 MHz, 2ms:8ms, 5min) stimulated collagen and non- 
collagenous protein synthesis, intensities of 0.5-2 W cm -2 inhibited these. The observed 
protein synthesis stimulation was attributed to osteoblastic activity. Yang et al. (1996) 
found a statistically significant increase in mechanical strength in fractured rat femurs 
at 0.05 W cm -2 (I , 0.5 MHz) but not at 0.1 W cm -2 . They noted a shift in the expression of 
genes associated with cartilage formation in the treated bones. Aggrecan gene expression 
was higher than control values on Day 7, but lower than control on Day 21. Wang et al. 
(1994) found ultrasonically accelerated fracture repair at 21 days in a rat femoral model, 
but only at 1.5 MHz (I 0.03 W cnr 2 , 200 ps: 800 ps) and not at 0.5 MHz. 



Ultrasound 
can accelerate 
bone repair 



95 



7 Bio-effects — cells and tissues 



7.3.2 Effects on soft tissues 

Ultrastructural changes resulting from ultrasonic exposure of intact soft tissues 
have been described above. Extracellular membranes, lysosomes and mitochondria 
are the cellular sites most readily affected. Ultrasonic exposure of soft-tissue 
wounds can accelerate healing (Dyson et al, 1968; Dyson, 1990; Young and Dyson, 
1990; De Deyne and Kirsch-Volders, 1995). This is probably due to the stimulation of 
protein synthesis. 



Ultrasound 
may stimulate 
wound healing 



Intensities above 1 W cm 2 have been shown to produce a variety of effects in smooth 
muscle. These include alteration of contractile activity (Talbert, 1975; ter Haar 
et al., 1978) and inhibition of action potentials (Hu et al., 1978). These effects are 
thought to be due to acoustic streaming mediated alteration of ion transport across 
cell membranes. 



7.3.3 Effects on blood components and vasculature 



Ultrasound may 
damage platelets 



The most fragile component of the vasculature is the platelet population. Damage to 
platelets is important because it carries with it the associated risk of thrombus formation. 
Williams (1974) and Miller et al. (1979) have shown that platelet damage may be induced 
in vitro by the shear stresses associated with an ultrasonic exposure, at levels lower than 
those needed to produce damage to erythrocytes. In the presence of stable bubbles, 
platelets in suspension in vitro may be damaged at spatial average intensities as low as 
0.8 W cm- 2 (Miller et al, 1979). 



Erythrocytes 
are resistant 
to ultrasound 
damage 



In contrast, it appears that erythrocytes are resistant to ultrasonically induced damage. 
In the presence of collapse (inertial cavitation) haemolysis has been observed (Rooney, 1970; 
Williams et al., 1970; Wong and Watmough, 1980). ATP may be released at lower spatial 
average intensities in the presence of non-inertial cavitation (Williams and Miller, 1980). 



It is difficult 
to produce 
cavitation in 
whole blood 



Whole blood in vivo is continuously filtered of impurities, and so is not rich in potential 
cavitation nuclei. It is therefore very difficult to induce cavitation in whole blood. 
However, Brayman et al. (1996a,b) have demonstrated that it may occur if the acoustic 
pressure is sufficiently high (~17MPa). The addition of gas-filled contrast agents to 
whole or diluted blood may reduce the acoustic pressure thresholds for the production 
of cavitation to as low as UMPa (Miller and Thomas, 1995; 1996; Brayman et al., 1995; 
Ivey et al, 1995). However, this is still significantly higher than the pressures found in 
commercial diagnostic scanners. 

In vivo attempts to detect damage to blood components has largely proved negative 
(Williams et al, 1977; Deng et al, 1996; Dalecki et al, 1997; Poliachik et al, 1999). This 
is perhaps unsurprising since it is to be expected that only a small volume of cells is 
likely to be affected, and any that were would rapidly become diluted by normal cells 
flowing into the area. Dalecki et al. (1997) were able to detect a clinically insignificant 
level of haemolysis (<4%) following exposure of mouse blood through the chest wall. At 
2.35 MHz only 0.46% haemolysis was detected for a pressure amplitude of lOMPa. 

96 



Bio-effects — cells and tissues 7 

Erythrocyte extravasation has been observed following ultrasound exposure of mammalian 
lungs. This is the subject of extensive review in Chapter 5. 

Haemorrhage has been observed close to foetal bone (1.2 MHz; peak positive pressure 4 
MPa, peak rarefactional pressure 2.5 MPa) (Dalecki et al., 1999). This has been attributed 
to the relative motion between partially ossified bones and surrounding tissues, which 
may result in damage to the fragile foetal blood vessels. A thermal mechanism has also 
been suggested (Bigelow et al, 2007). 

Damage to the plasma membrane of the luminal aspect of endothelial cells of the blood 
vessels in the chick embryo and in mouse uterus have been seen following exposure to 
ultrasonic standing waves in vivo (Dyson et al, 1974; ter Haar et al., 1979). 

7.3.4 Consequences of ultrasonic exposure of embryos in utero 

The most important question to be considered when the safety of obstetric ultrasound is 
under consideration is whether ultrasound has any deleterious effect on the embryo when 
irradiated in utero. The literature on this topic has been extensively reviewed (Church 
and Miller, 2007; NCRP, 2002; Miller et al, 2002; AGNIR, 2010). This question has been 
addressed through studies into gross teratogenetic effects, developmental changes and 
genetic effects. The majority of these studies have been carried out in rats and mice. While 
these yield some interesting results, there are a number of limitations to these studies. In 
particular, the acoustic field is difficult to quantify in these small animals, and in most 
cases, the beam may have exposed a large fraction of the animal, with the consequent 
possibility of whole-body heating (Miller et al., 2002; Edwards et al., 2003). 

Raised maternal or foetal temperatures can result in a spectrum of adverse outcomes 
that affect many developing tissues, but especially the developing brain and nervous 
system. While the likelihood is low, it is possible that pulsed ultrasound could affect 
the integrity of maternal and developing tissues through non-thermal interactions, 
and especially by cavitational mechanisms when ultrasound contrast agents are present 
(Abramowicz, 2005). 

Whole embryo culture is sometimes used as a model system for the study of external 
influences on embryonic development. Ramnarine et al. (1998) exposed rat embryos in 
culture to ultrasound with a range of intensity levels at frequencies between 1 MHz and 
4 MHz for 30min. No significant effects were found for spatial-peak temporal-average 
intensities below 4 Won 2 or negative pressures below 1.9 MPa. These levels are higher 
than used in diagnostic ultrasound machines (see Chapter 3). At higher exposure levels 
some gross morphological abnormalities such as cephalocaudal flexion and abnormal 
head development were observed. Both thermal and cavitational mechanisms were 
implicated in these findings. A maximum temperature rise of 3.5 °C was observed 
within the sample holder. Bubbles were produced during the higher ultrasound 
exposures, indicating that cavitation had occurred. There is no good evidence that 
cavitation occurs in vivo at diagnostic ultrasound exposure levels. Moreover, 30min is a 
longer exposure time than would be expected during clinical examinations (ter Haar, 2008). 

97 



Ultrasound 
effects have 
been reported in 
embryo culture 



7 Bio-effects — cells and tissues 



There have been a large number of studies of the effect of ultrasound on embryonic and 
foetal development following exposure in utero. In general, no abnormalities have been 
observed in the absence of a temperature rise to levels above 41.5 °C (Shoji et al, 1975; 
Hara et al, 1977; Hara, 1980). Lele (1979), Edwards (1986, 1993) and Abramowicz et al. 
(2008) have shown that uterine hyperthermia can lead to a number of teratological effects, 
foetal resorption and growth retardation. 



No effects have 
been seen in 
embryos or 
foetuses of 
experimental 
animals exposed 
to ultrasound in 
utero 



The effect of in utero ultrasound exposure on neuronal migration in mice on day 16 of 
gestation has been studied (Ang et al., 2006). There was no difference in brain size or gross 
cortical architecture when studied 10 days afterbirth, but there was a statistically significant 
dose dependent difference in neuronal dispersion in animals that had been exposed to 
ultrasound for 30min or more, with -4% more neurons in the experimental group of 
animals remaining in the deeper neuronal layers after 60min of exposure. In control 
animals all neurons had reached the superficial layers. A number of important factors must 
be considered when extrapolating these results to the human. For experimental reasons, 
the pregnant females required restraint during exposure. This, alone, influences neuronal 
migration, as was shown by the increased dispersion in the sham control animals. There 
was very little ultrasound attenuation in the tissue path overlying the foetal mice, leading 
to much higher intensities than might be experienced by the human foetus. The whole 
foetal mouse was exposed, whereas only a small proportion of the human foetus would be 
using these probes. It is therefore difficult to extrapolate these results to human exposures, 
and their functional significance is difficult to judge, since the neurons involved may not 
persist. 



There have 
been reports 
that ultrasound 
exposures may 
lead to reduced 
birth weight 



A number of reports have suggested that ultrasonic irradiation in utero may result in 
foetal weight reduction (O'Brien, 1976; Stolzenberg et al, 1980; Tachibana et al, 1977). 
Barnett and Williams (1990) concluded that the adverse effects seen were caused by 
changes in maternal physiology produced largely by heating of maternal organs such 
as the bladder and spine. 



The majority of the studies reported in the literature have been conducted at the potentially 
most sensitive stage of organogenesis (about 8 days in the mouse, 9 days in the rat). 
Despite this, no adverse effects have been seen at exposure levels used diagnostically. 
It therefore seems unlikely that current obstetric diagnostic ultrasound practice will 
lead to teratogenic or developmental changes. Available human epidemiological studies 
reinforce this finding (see Chapter 9). 



7.4 Conclusion 

While a number of ultrasonically induced changes inbiological systems have been reported, 
it is important that they be assessed in an informed manner. It must be remembered that 
the mechanisms by which effects may be produced in vitro may not carry the same weight 
as those that operate in vivo. In addition, the difficulties in scaling ultrasonic beams to an 
appropriate size for small animal exposures mean that a proportionately much larger 
fraction of, for example, a rat uterus, is exposed by a stationary beam than is the case for 
a human foetus. 

98 



Bio-effects — cells and tissues 7 



Nevertheless, useful information can be gleaned from the existing literature, and 
while there is currently no good biological evidence which suggests that ultrasound for 
obstetric examinations should be withheld, constant vigilance and continued, targeted 
research is essential. 

Acknowledgement 

This chapter is a revised version of Chapter 7 in the second edition. The contribution of 
Stan Barnett to that chapter is acknowledged. 

References 

Abramowicz JS. 2005. Ultrasonographic contrast media: has the time come in obstetrics 
and gynecology? / Ultrasound Med, 24, 517-531. 

Abramowicz JS, Barnett SB, Duck FA, Edmonds PD, Hynynen KH, Ziskin MC. 2008. Fetal 
thermal effects of diagnostic ultrasound. / Ultrasound Med, 27, 541-559. 
AGNIR. 2010. Health Effects of Exposure to Ultrasound & Infrasound Advisory Group on 
Non-ionising Radiation. London, UK: Health Protection Agency. 

Ang Jr ES, Gluncic V, Duque A, Schafer ME, Rakic P. 2006. Prenatal exposure to ultrasound 
waves impacts neuronal migration in mice. Proc Natl Acad Sci USA, 103, 12903-12910. 
Barnett SB, Williams AR. 1990. Identification of possible mechanisms responsible for fetal 
weight reduction in mice following exposure. Ultrasonics, 28, 159-165. 
Bigelow TA, Miller RJ, Blue JP, O'Brien Jr WD. 2007. Haemorrhage near fetal rat bone 
exposed to pulsed ultrasound. Ultrasound Med Biol, 33, 311-317. 

Bleaney BI, Blackbourne P, Kirkley J. 1972. Resistance of CHLF hamster cells to ultrasonic 
radiation of 1.5 MHz frequency. Br ] Radiol, 45, 354-357. 

Bosward KL, Barnett SB, Wood AKW, Edwards MJ, Kossoff G. 1993. Heating of guinea- 
pig fetal brain during exposure to pulsed ultrasound. Ultrasound Med Biol, 19, 415-424. 
Brayman AA, Doida Y, Miller MW. 1992. Apparent contribution of respiratory gas 
exchange to the in vitro "cell density effect" in ultrasonic cell lysis. Ultrasound Med Biol, 
18, 701-714. 

Brayman AA, Azadniv M, Makin IRS, Miller MW, Carstensen EL, Child SZ, et al. 1995. 

Effect of a stabilised microbubble contrast agent on haemolysis of human erythrocytes 

exposed to high intensity pulsed ultrasound. Echocardiography, 12, 13-21. 

Brayman AA, Azadniv M, Cox C, Miller MW. 1996a. Haemolysis of Albunex-supplemented, 

40% haematocrit human erythrocytes in vitro by 1 MHz pulsed ultrasound: acoustic 

pressure and pulse length dependence. Ultrasound Med Biol, 22, 927-938. 

Brayman AA, Church CC, Miller MW. 1996b. Re-evaluation of the concept that high cell 

concentrations "protect" cells in vitro from ultrasonically induced lysis. Ultrasound Med 

Biol, 22, 497-514. 

Carstensen EL, Child SZ, Norton S, Nyborg WL. 1990. Ultrasonic heating of the skull. 
JAcoust SocAm, 87, 1310-1317. 

Carstensen EL, Kelly P, Church CC, Brayman AA, Child SZ, Raeman CH, et al. 1993. Lysis 
of erythrocytes by exposure to CW ultrasound. Ultrasound Med Biol, 19, 147-165. 
Chang WH, Sun JS, Chang SP, Lin JC. 2002. Study of thermal effects of ultrasound 
stimulation on fracture healing. Bioelectromagnetics, 23, 256-263. 

99 



7 Bio-effects — cells and tissues 



Chapman IV. 1974. The effect of ultrasound on the potassium contents of rat thymocytes 
in vitro. Br ] Radiol, 47, 411-415. 

Church CC, Miller MW. 2007. Quantification of risk from fetal exposure to diagnostic 
ultrasound. Prog Biophys Mol Biol, 93, 331-353. 

Claes L, Willie B. 2007. The enhancement of bone regeneration by ultrasound. Prog Biophys 
Mol Biol, 93, 384-398. 

Clarke PR, Hill CR. 1969. Biological action of ultrasound in relation to the cell cycle. Exp 
Cell Res, 58, 443. 

Dalecki D, Raeman CH, Child SZ, Cox C, Francis CW, Meltzer RS, et al. 1997. Haemolysis 
in vivo from exposure to pulsed ultrasound. Ultrasound Med Biol, 23, 307-313. 
Dalecki D, Child SZ, Raeman CH, Cox C. 1999. Haemorrhage in murine fetuses exposed 
to pulsed ultrasound. Ultrasound Med Biol, 25, 1139-1144. 

De Deyne PG, Kirsch-Volders M. 1995. In vitro effects of therapeutic ultrasound on the 
nucleus of human fibroblasts. Phys Ther, 75, 629-634. 

Deckers R, Moonen C. 2010. Ultrasound triggered, image guided, local drug delivery. 
/ Control Release, 148, 25-33. 

Deng CX, Xu Q, Apfel RE, Holland CK. 1996. In vitro measurements of inertial cavitation 
thresholds in human blood. Ultrasound Med Biol, 22, 939-948. 

Duarte LR. 1983. The stimulation of bone growth by ultrasound. Arch Orthop Trauma 
Surg, 101, 153-159. 

Dyson M. 1990. Role of ultrasound in wound healing. In Wound Healing: Alternatives in 
Management, Kloth LC, McCulloch JM, Feedar JA (editors). Philadelphia, PA: FA Davis, 
pp. 259-285. 

Dyson M, Brookes M. 1983. Stimulation of bone repair by ultrasound. In Ultrasound '82, 
Lerski RA, Morley P (editors). Oxford, UK: Pergamon Press, pp. 61-66. 
Dyson M, Pond JB, Joseph J, Warwick R. 1968. The stimulation of tissue regeneration by 
means of ultrasound. Clin Sci, 35, 273-295. 

Dyson M, Pond JB, Woodward B, Broadbent J. 1974. The production of blood cell stasis 
and endothelial damage in the blood vessels of chick embryos treated with ultrasound in 
a stationary wave field. Ultrasound Med Biol, 1, 133-148. 

Dyson M, Smalley D. 1983. Effects of ultrasound on wound contraction. In Ultrasound 
Interactions in Biology and Medicine, Millner R, Corbet U (editors). New York, NY: 
Plenum Press, p. 151. 

Duggan PM, Liggins GC, Barnett SB. 1995. Ultrasonic heating of the brain of the fetal 
sheep in utero. Ultrasound Med Biol, 21, 553-560. 

Dvorak M, Hrazdira I. 1966. Changes in the ultrastructure of bone marrow cells in rats 

following exposure to ultrasound. Z Mikrosk Anat Forsch, 4, 451-460. 

Edwards MJ. 1986. Hyperthermia as a teratogen: a review of experimental studies and 

their clinical significance. Teratog Carcinog Mutagen, 6, 563-582. 

Edwards MJ. 1993. Hyperthermia and birth defects. Cornell Vet, 83, 1-7. 

Edwards MJ, Saunders RD, Shiota K. 2003. Effects of heat on embryos and foetuses. 

Int ] Hyperthermia, 19, 295-324. 

EFSUMB. 1994. European Federation of Societies for Ultrasound in Medicine & Biology. 
Tutorial paper: genetic effects. Eur ] Ultrasound, 1, 91-92. 

Ellwart JW, Brettel H, Kober LO. 1988. Cell membrane damage by ultrasound at different 
cell concentrations. Ultrasound Med Biol, 14, 43-50. 

100 



Bio-effects — cells and tissues 7 



Feril Jr LB, Kondo T. 2004. Biological effects of low intensity therapeutic ultrasound in 
vitro: the potentials for therapy and the implications on safety of diagnostic ultrasound. 
Int Congress Ser, 1274, 133-140. 

Fyfe M, Chahl LA. 1982. Mast cell degranulation: a possible mechanism of action of 
therapeutic ultrasound. Ultrasound Med Biol, 8(Suppl. 1), 62. 

Gebhart E. 1981. Sister chromatid exchange (SCE) and structural chromosome aberration 
in mutagenicity testing. Hum Genet, 58, 235-254. 

ter Haar G. 2007. Therapeutic applications of ultrasound. Prog Biophys Mol Biol, 93, 
111-129. 

ter Haar GR. 2008. Results of a survey of exposure conditions used in ultrasound scans in 
the UK, February 2007. Ultrasound, 16, 110-113. 

ter Haar GR, Dyson M, Talbert D. 1978. Ultrasonically induced contractions in mouse 
uterine smooth muscle in vivo. Ultrasonics, 16, 275-276. 

ter Haar GR, Dyson M, Smith SP. 1979. Ultrastructure changes in the mouse uterus 
brought about by ultrasonic irradiation at therapeutic intensities in standing wave fields. 
Ultrasound Med Biol, 5, 167-179. 

ter Haar GR, Stratford LJ, Hill CR. 1980. Ultrasonic irradiation of mammalian cells in vitro 
at hyperthermic temperatures. Br J Radiol, 53, 784-789. 

Hallow DM, Mahajan AD, McCutchen TE, Prausnitz MR. 2006. Measurement and 
correlation of acoustic cavitation with cellular bioeffects. Ultrasound Med Biol, 32, 1111-1122. 
Hara K, Minoura S, Okai T, Sakamoto S. 1977. Symposium on recent studies in the 
safety of diagnostic ultrasound, safety of ultrasonics on organism. Jpn J Med Ultrason, 
4, 256-258. 

Hara K. 1980. Effect of ultrasonic irradiation on chromosomes, cell division and developing 
embryos. Acta Obstet Gynecol Jpn, 32, 6148. 

Harvey W, Dyson M, Pond JB, Grahame R. 1975. The in vitro stimulation of protein 
synthesis in human fibroblasts by therapeutic levels of ultrasound. In Ultrasonics in 
Medicine, Kazner E, de Vlieger M, Muller HR, McCready VR (editors). Excerpta Medica, 
International Congress, Series No. 363. Amsterdam, the Netherlands: Excerpta Medica, 
pp. 10-21. 

Heckman JD, Ryaby JP, McCabe J, Frey JF, Kilcoyne RF. 1994. Acceleration of tibial fracture 
healing by non-invasive, low intensity pulsed ultrasound. / Bone Joint Surg, 76, 26-34. 
Horder MM, Barnett SB, Vella GJ, Edwards MJ, Wood AKW. 1998. In vivo heating of the 
guinea-pig fetal brain by pulsed ultrasound and estimates of thermal index. Ultrasound 
Med Biol, 24, 1467-1474. 

Hrazdira L. 1970. Changes in cell ultrastructure under direct and indirect action of 
ultrasound. In Ultrasonographia Medica, Bock J, Ossoinig K (editors). Vienna, Austria: 
Vienna Academy of Medicine, pp. 457-463. 

Hu JH, Taylor JD, Press HC, White JE. 1978. Ultrasonic effects on mammalian interstitial 
muscle membrane. Aviat Space Environ Med, 49, 607-609. 

Ivey JA, Gardner EA, Fowlkes JB, Rubin JM, Carson PL. 1995. Acoustic generation of 
intra-arterial contrast boluses. Ultrasound Med Biol, 21, 757-767. 

Joshi GP, Hill CR, Forrester JA. 1973. Mode of action of ultrasound on the surface charge 
of mammalian cells. Ultrasound Med Biol, 1, 45^8. 

Kaufman GE, Miller MW, Griffiths TD, Ciaravino V. 1977. Lysis and viability of cultured 
mammalian cells exposed to 1 MHz ultrasound. Ultrasound Med Biol, 3, 21-25. 

101 



7 Bio-effects — cells and tissues 



Kunze-Muhl E. 1981. Observations on the effect of X-ray alone and in combination with 
ultrasound on human chromosomes. Hum Genet, 57, 257-260. 

Lai CY, Wu CH, Chen CC, Li PC. 2007. Quantitative relations of acoustic inertial cavitation 

with sonoporation and cell viability. Ultrasound Med Biol, 32, 1931-1941. 

Latt SA, Schreck RR. 1980. Sister chromatid exchange analysis. Am ] Hum Genet, 32, 297. 

Li GC, Hahn GM, Tolmach LJ. 1977. Cellular inactivation by ultrasound. Nature, 267, 

163-165. 

Liebeskind D, Bases R, Mendez F, Elequin F, Koenigsberg M. 1979. Sister chromatid 
exchanges in human lymphocytes after exposure to diagnostic ultrasound. Science, 205, 
1274-1275. 

Liebeskind D, Padawer J, Wolley R, Bases R. 1982. Diagnostic ultrasound: time lapse and 
transmission electron microscopic studies of cells insonated in vitro. Br } Cancer Suppl, 
45(Suppl. V), 176-186. 

Lele PP. 1979. Safety and potential hazards in the current applications of ultrasound in 
obstetrics and gynaecology. Ultrasound Med Biol, 5, 307-320. 

Miller DL, Thomas RM. 1995. Ultrasound contrast agents nucleate inertial cavitation in 
vitro. Ultrasound Med Biol, 21, 1059-1065. 

Miller DL, Thomas RM. 1996. Contrast agent gas bodies enhance haemolysis induced by 
lithotriptor shock waves and high intensity focused ultrasound in whole blood. Ultrasound 
Med Biol, 22, 1089-1095. 

Miller MW, Nyborg WL, Dewey WC, Edwards MJ, Abramowicz JS, Brayman AA. 2002. 
Hyperthermic teratogenicity, thermal dose and diagnostic ultrasound during pregnancy: 
implications of new standards on tissue heating. Int } Hyperther, 8, 361-384. 
Mortimer AJ, Dyson M. 1988. The effect of therapeutic ultrasound on calcium uptake in 
fibroblasts. Ultrasound Med Biol, 14, 499-506. 

Morton KI, ter Haar GR, Stratford LJ, Hill CR. 1982. The role of cavitation in the interaction 
of ultrasound with V79 Chinese Hamster cells in vitro. Br } Cancer, 45, 147-150. 
Morton KI, ter Haar GR, Stratford LJ, Hill CR. 1983. Subharmonic emission as an indicator 
of ultrasonically induced biological damage. Ultrasound Med Biol, 9, 629-633. 
Miller DL. 2007. Overview of experimental studies of biological effects of medial ultrasound 
caused by gas body activation and inertial cavitation. Prog Biophys Mol Biol, 93, 414-430. 
Miller DL, Nyberg WL, Whitcomb CC. 1979. Platelet aggregation induced by ultrasound 
under specialized conditions in vitro. Science, 205, 505-507. 

Mummery CL. 1978. Effect of ultrasound on fibroblasts in vitro. Ph.D. thesis, University 
of London. 

NCRP. 2002. Exposure Criteria for Medical Diagnostic Ultrasound: II Criteria Based on all 
Known Mechanisms. NRCP Report no. 140. Bethesda, MD: National Council on Radiation 
Protection and Measurements. 

O'Brien WD. 1976. Ultrasonically induced fetal weight reduction in mice. In Ultrasound 
in Medicine, White DN (editor). New York, NY: Plenum Press, pp. 531-532. 
Pilla AA, Mont MA, Nasser PR, Khan SA, Figueiredo M, Kaufman JJ, et al. 1990. Non- 
invasive low intensity pulsed ultrasound accelerates bone healing in the rabbit. / Orthop 
Trauma, 4, 246-253. 

Poliachik SL, Chandler WL, Mourad PD, Bailey MR, Bloch S, Cleveland RO, et al. 1999. 
Effect of high intensity focused ultrasound on whole blood with and without microbubble 
contrast agent. Ultrasound Med Biol, 25, 991-998. 

102 



Bio-effects — cells and tissues 7 



Ramnarine KV, Nassiri DK, McCarthy A, Brown NA. 1998. Effects of pulsed ultrasound 
on embryonic development: an in vitro study. Ultrasound Med Biol, 24, 575-585. 
Reher P, Elbeshir E-NI, Harvey W, Meghji S, Harris M. 1997. The stimulation of bone 
formation in vitro by therapeutic ultrasound. Ultrasound Med Biol, 23, 1251-1258. 
Rooney JA. 1970. Hemolysis near an ultrasonically pulsating gas bubble. Science, 169, 
869-871. 

Rott H-D. 1981. Zur Frage der Schadigungsmoglichkeit durch diagnostis-chen Ultraschall. 

Ultraschall Med, 2, 56. 

Shoji R, Murackami U, Shimizu T. 1975. Influence of low intensity ultrasonic irradiation 
on prenatal development of two inbred mouse strains. Teratology, 12, 227-232. 
Stolzenberg SJ, Torbit CA, Pryor GT, Edmonds PD. 1980. Toxicity of ultrasound in mice: 
neonatal studies. Radiat Environ Biophys, 18, 37-44. 

Tachibana M, Tachibana Y, Suzuki M. 1977. The present status of the safety of ultrasonic 
diagnosis in the area of obstetrics — the effect of ultrasound irradiation on pregnant mice 
as indicated in their fetuses. Jpn } Med Ultrason, 4, 279-283. 

Talbert DG. 1975. Spontaneous smooth muscle activity as a means of detecting biological 
effects of ultrasound. In Proceedings of Ultrasonics International. Guildford, UK: IPC 
Science & Technology Press, pp. 279-284. 

Taylor KJW, Newman DL. 1972. Electrophoretic mobility of Ehrlich suspensions exposed 
to ultrasound of varying parameters. Phys Med Biol, 17, 270-276. 

Taylor KJW, Pond JB. 1972. Primary sites of ultrasonic damage on cell systems. In 
Interaction of Ultrasound and Biological Tissues, Reid M, Sikov MR (editors). Washington, 
DC: DHEW Publication No. (FDA)73-8008. 

Thacker J. 1973. The possibility of genetic hazard from ultrasonic radiation. Curr Top 
Radiat Res Q, 8, 235-258. 

Tsai C-L, Chang WH, Liu T-K. 1992. Preliminary studies of duration and intensity of 
ultrasonic treatments on fracture repair. Chin J Physiol, 35, 21-26. 

Wang SJ, Lewallen DG, Bolander ME, Chao EYS, Ilstrup DM, Greenleaf JF. 1994. Low 
intensity ultrasound treatment increases strength in a rat femoral fracture model. 
/ Orthop Res, 12, 40-47. 

Webster DF, Dyson M, Harvey W. 1979. Ultrasonically induced stimulation of collagen 
synthesis in vivo. In Proceedings of the 4th European Symposium on Ultrasound in Biology and 
Medicine, 1, pp. 135-140. 

Webster DF, Harvey W, Dyson M, Pond JB. 1980. The role of ultrasound-induced 
cavitation in the in vitro stimulation of collagen synthesis in human fibroblasts. Ultrasonics, 
18, 33-37. 

Webster DF, Pond JB, Dyson M, Harvey W. 1978. The role of cavitation in the in vitro 
stimulation of protein synthesis in human fibroblasts by ultrasound. Ultrasound Med Biol, 
4,343-351. 

Williams AR, Hughes DE, Nyborg WL. 1970. Hemolysis near a transversely oscillating 
wire. Science, 169, 871-873. 

Williams AR. 1974. Release of serotonin from human platelets by acoustic microstreaming. 
JAcoust SocAm, 56, 1640-1643. 

Williams AR, Chater BV, Sanderson JH, Taberner DA, May SJ, Allen KA, et al. 1977. Beta- 
thromboglobulin release from human platelets after in vivo ultrasound irradiation. Eancet, 
2, 931-932. 

103 



7 Bio-effects — cells and tissues 



Williams AR, Miller DL. 1980. Photometric detection of ATP release from human 
erythrocytes exposed to ultrasonically activated gas-filled pores. Ultrasound Med Biol, 6, 
251-256. 

Wong YS, Watmough DJ. 1980. Hemolysis of red blood cells in vitro and in vivo caused by 
therapeutic ultrasound at 0.75 MHz. In Proceedings of Ultrasound Interaction in Biology and 
Medicine Symposium, Reinhardsbrunn, GDR, 10-14 November. Paper C-14. 
Yang KH, Parvizi J, Wang SJ, Lewallen DG, Kinnick RR, Greenleaf JF, et al. 1996. Exposure 
to low intensity ultrasound increases the aggrecan gene expression in a rat femur fracture 
model. / Orthop Res, 14, 802-809. 

Young SR, Dyson M. 1990. Effect of therapeutic ultrasound on the healing of full-thickness 
excised skin lesions. Ultrasonics, 28, 175-180. 



104 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 8 

The safe use of contrast-enhanced 

diagnostic ultrasound 

Douglas L. Miller 

Department of Radiology, University of Michigan, Ann Arbor, Ml, USA 



Summary 

• Ultrasound contrast agents are complex drugs consisting of suspended gas bodies. 

• Post-marketing surveillance has identified adverse events, which led to the inclusion 
of new warnings and contraindications in package inserts. 

• Ultrasonic activation of the gas bodies produces strong echoes, and concentrates 
mechanical activity in their vicinity, which can lead to localized biological effects. 

• Reported bio-effects include subcellular lesions, cell lysis, capillary leakage and 
haemorrhage, lethal injury of cardiomyocytes, cardiac arrhythmia, renal tubular 
obstruction and haematuria. 

• The current mechanical index does not provide actual dosimetric information 
for biological effects when gas-body based contrast agents are used in diagnostic 
ultrasound. 

• The optimum risk-to-benefit ratio may be obtained for the patient through the diligent 
application of the contraindications, warnings and usage instructions for ultrasound 
contrast agents. 



8.1 Introduction 



The concept of contrast enhancement by an external agent originated early in the evolution 
of diagnostic imaging by ultrasound. Blood-filled regions often appear empty in an 
ultrasound image, and provision of material (contrast agent), which brings the image 
brightness above the normal background of blood or tissue, can yield new diagnostic 
information. Beginning with observations of contrast from saline injections (Gramiak, 
1997), the development of ultrasound contrast agents has progressed to an expanding 
list of commercially-developed agents which are approved for clinical use. Ultrasound 
contrast agents are unique in several ways. Since ultrasound images involve echoes, a 
passive absorptive agent (analogous to the iodinated agents used in X-ray angiography) 
would be of little value. The ideal contrast agents for ultrasound are suspensions of 

105 



Modern 
ultrasound 
contrast 
agents consist 
of suspensions 
of stabilized 
microbubbles 
engineered to 
persist when in 
the circulation 



8 The safe use of contrast-enhanced diagnostic ultrasound 



microbubbles that can circulate through the body and provide strong pulse echoes. 
Contrast echocardiography has been shown to be useful in cardiology both for left 
ventricular boarder delineation and for assessing perfusion of the myocardium (Porter 
and Xie, 2010). In addition, contrast-enhanced diagnostic ultrasound (CEDUS) is utilized 
in abdominal organs, particularly for evaluation of liver masses (Cosgrove, 2006; Wilson 
and Burns, 2010). In this review of safety issues, the advantages and benefits of CEDUS 
will not be described in detail; the reader is referred to several comprehensive books 
(Nanda et al, 1997; Thomsen et al, 1999; Goldberg et al, 2001) and numerous reviews on 
specific applications. 



Safety issues for 
the clinical use 
of ultrasound 
contrast materials 
include adverse 
reactions to 
the drugs and 
biological effects 
associated 
with ultrasonic 
cavitation 



Since the first review of the safe use of ultrasound contrast agents (Miller, 2000), the 
progress and adoption of CEDUS has been negatively impacted by safety issues, which 
are the topic of this update. Ultrasound contrast agents present two distinct types of 
risks. First, these agents have many of the same potential risk factors as other injectable 
agents, such as adverse reactions, which have been of concern for regulatory agencies. 
Second, diagnostic ultrasound pulses not only activate the gas bodies, producing strong 
and unique echoes for imaging, but can also nucleate ultrasonic cavitation, a potential 
mechanism for bio-effects. Without the addition of such exogenous cavitation nuclei, 
cavitation thresholds are expected to be high (Church, 2002; Carstensen et al., 2000). 
Thus, the injection of gas-body based contrast agents into the circulation introduces an 
entirely novel risk of bio-effects for diagnostic ultrasound. Several authoritative reviews 
of bio-effects considerations for CEDUS have been conducted by the National Council 
on Radiation Protection (NCPJ 3 , 2002), the World Federation of Ultrasound in Medicine 
and Biology (Barnett et al., 2007) and the American Institute of Ultrasound in Medicine 
(Miller et al., 2008a) among others. Unfortunately, the research has not been exhaustive 
or conclusive, leaving many uncertainties. The purpose of this update is to review recent 
developments in ultrasound contrast agents, their ultrasonic behaviour, pharmacological 
safety issues, bio-effects research and guidance for the safe use of CEDUS. 



8.2 Ultrasound contrast agents and how they work 

8.2.1 Gas-body design 

Advanced Some popular commercial agents, which have been approved for clinical use, are listed 

contrast agents - n j^jg g j j n addition, other agents are under development and in clinical trails, which 
use low solubility ° r 

gases and a ma y be approved in the near future. The two agents Echovist and Levovist (Schering 

stabilizing skin to AG, Berlin) consist of dry galactose particles which are manually reconstituted for use 

(Schlief et al., 1993). The ultrasonic attenuation produced by Levovist peaks at about 

2-3 MHz, indicating a typical bubble size of 2.7 um. Albunex (Molecular Biosystems, Inc. 

or Mallinckrodt, Inc.) was created by sonication of a solution of human serum albumin, 

which produced numerous gas bodies with stable shells of denatured albumin about 

15 nm in thickness (Christiansen et al., 1994). Advanced agents have been invented to 

improve persistence in the circulation. Optison (GE Healthcare, Princeton, NJ, USA) 

is related to the earlier product Albunex, but contains perflutren, a gas which is much 

less soluble in aqueous media than air (Killam and Dittrich, 1997). The product contains 

about 5-8 x 10 s per ml of gas bodies with a mean diameter range of 3.0-4.5 um, which 

106 



The safe use of contrast-enhanced diagnostic ultrasound 8 



Table 8.1. A partial listing of ultrasound contrast agents, which have been approved for 
clinical use. 



Agent Stabilization Gas 

designation 



Associated 
company 



Exemplary 
citation 



Echovist 



Albunex 



Levovist 



Optison 



Definity 



Sonovue 



Galactose 



Albumin 



Galactose 



Albumin 



Lipids 



Lipids 



Imagent US Surfactants 



Air 



Air 



Air 



Perflutren 



Sulphur 
hexafluoride 

Perflexane 



Schering AG 
Mallinckrodt 
Schering AG 
Mallinckrodt 



Octafluoropropane ImaRx Pharm. 



Bracco 
Diagnostics 

Alliance Pharm. 



Schlief et al., 
1993 

Bleeker et ah, 
1990 

Schlief et al., 
1993 

Killam and 
Dittrich, 1997 

Unger et al., 
1997 

Schneider 
et al, 1995 

Mattrey and 
Pelura, 1997 



are stable under refrigeration for several years. Definity (Lantheus Medical Imaging, 
N Billerica, MA, USA) is prepared by shaking a 2 ml vial containing clear liquid and 
octafluoropropane gas, which produces a creamy suspension containing 1.2 *10 10 
lipid-stabilized microbubbles per ml with a 1.1-3.3 pm mean diameter range. Imagent 
(Perflexane Lipid Microsphere, Alliance Pharmaceutical, San Diego, CA, USA) is 
reconstituted with 10ml of water, which results in about 10 9 per ml of lipid-stabilized 
microbubbles of 6 pm volume-weighted mean diameter. Sonovue (Bracco International, 
Amsterdam, The Netherlands) is reconstituted from a dry powder with 5 ml of saline, 
which yields a suspension of 2-5 x 10 8 lipid-stabilized microbubbles per ml of 1-10 pm 
diameter with sulphur hexafluoride gas. 



8.2.2 Physics of ultrasound contrast agents 

Commercial ultrasound contrast agents consist of gas bodies in a carrier medium. Gas 
bodies interact with a low-amplitude acoustic pressure wave by pulsating, much like a 
simple resonant oscillator. The size range required for passage through the circulation, 
less than about 4 pm radius, fortuitously coincides with the range of strongly responding 
sizes of bubbles in the MHz frequency range of diagnostic ultrasound. Although the 
stabilizing shells or skins reduce the pulsation relative to free gaseous microbubbles, 
measurements of scattering have shown that these gas bodies are more effective, by 
orders of magnitude, in producing echoes than are cells of the blood. 



Contrast gas 
bodies less than 
about 4 mm 
radius can pass 
through the 
capillary bed: 
they also provide 
strong echoes 
at diagnostic 
ultrasound 
frequencies 



At amplitudes relevant to diagnostic imaging, the scattered ultrasound includes the 
second harmonic, which can be exploited in harmonic imaging modes of ultrasound 
scanners to produce stronger contrast between the agents and tissue (Krishna and 
Newhouse, 1997). Another useful non-linear signal is the subharmonic, which may have 

107 



8 The safe use of contrast-enhanced diagnostic ultrasound 



Microbubble 
pulsation is 
non-linear, which 
allows special 
contrast-specific 
imaging modes 
in CEDUS 



advantages over the second harmonic for imaging (Forsberg et ah, 2000). The non-linear 
behaviour can be exploited by using special imaging modes designed to separate the 
gas-body echoes from normal tissue echoes (Averkiou et ah, 2003). These modes can 
utilize relatively low power to image tissue perfusion with contrast agent in real-time 
(Cosgrove and Blomley, 2004). The actual tissue blood supply at the capillary level is 
revealed, which is not possible for non-contrast diagnostic ultrasound imaging. 



Contrast agents 
are easily 
destroyed by 
diagnostic 
ultrasound, 
eliminating 
the contrast- 
enhancement 
phenomenon 



8.2.3 Gas-body stability and nucleation of inertial cavitation 

Although the commercial gas-body contrast agents are stable in storage for extended 
periods, most are unstable during handling and use. The gas bodies are also susceptible 
to destruction by ultrasound, because the stabilizing films or shells are fragile (Stride 
and Saffari, 2003). At low pulse raref actional pressure amplitudes (RPAs), the shells 
can eff actively prevent expansion, resulting in a one-sided oscillation (Marmottant 
et ah, 2005). However, at higher amplitudes the gas bodies become unstable, and are 
effectively destroyed by the diagnostic ultrasound. The destruction of the gas bodies 
can be complex and depends strongly on the ultrasonic RPAs (Shi et ah, 2000; Chomas 
et ah, 2001). Basically, however, at modest RPAs the stabilization is lost, which leads to 
release of free gaseous microbubbles, followed by dissolution at the conclusion of the 
pulse (Porter et ah, 2006). 



Released 
microbubbles 
may act as 
nuclei for inertial 
cavitation, 
an important 
potential 
mechanism for 
biological effects 



At the time of destabilization, the liberated microbubble can exhibit a greatly increased 
echo amplitude. This process can also be viewed as the nucleation of cavitation activity. 
The possible initiation of inertial cavitation during CEDUS is important because 
cavitation is a well-known potential mechanism for biological effects of ultrasound 
(see Chapter 5). Secondary mechanisms for bio-effe cts arising from the cavity dynamics 
may include: volume pulsation, fragmentation, mechanical displacement and tearing 
of structures, capillary circumferential stress, formation of destructive liquid jets, 
acoustic microstreaming, transient hot-spots and creation of free radicals by the extreme 
temperatures achieved inside microbubbles upon inertial collapse (NCRP, 2002). 



Early in the consideration of diagnostic ultrasound safety, Apfel (1982) aid Flynn 
(1982) showed that cavitation was theoretically possible for diagnostic ultrasound. 
The thresholds for nucleation of cavitation in blood when ideal nuclei were present 
was analysed by Apfel and Holland (1991). The threshold ranged from about 0.4MPa 
at 1MHz to 1.1 MPa at 10 MHz, which are well within the capabilities of diagnostic 
ultrasound machines. 



Potential adverse 
drug reactions 
to ultrasound 
contrast agents 
have been 
detected in the 
sensitive porcine 
model of human 
imaging 



8.3 Pharmaceutical considerations and post-marketing 
surveillance 

8.3.1 Drug approval and regulation 

The clinical safety and efficacy of ultrasound contrast agents are evaluated by regulatory 
agencies before approval for use. These agents carry the general risks of bleeding and 
infection that pertain to any injection procedure. The agents are approved as injectable 

108 



The safe use of contrast-enhanced diagnostic ultrasound 8 



drugs, but are not intended to be pharmacologically active. After intravenous injection, most 
agents are depleted from the circulation by the lungs or by the reticuloendothelial system, 
with a potential for adverse reactions. In research, anaphylactoid reactions are rare in rats, 
but are frequent and severe in swine, which makes this animal a sensitive test subject for 
drug reactions (e.g. Szebeni et al., 2007). Such reactions are known to occur from ultrasound 
contrast agents, including Definity (Grauer et al., 1996), Albunex (Ostensen et al., 1992) and 
Sonovue (Bramos et al., 2008). In a recent study of renal bio-effects in swine (Miller et al., 
2010a, see below), the reactions to contrast agent infusion complicated the experiments with 
4 of 48 (8.3%) swine having life-threatening reactions. These reactions occurred even after 
brief pre- treatment with Diphenhydramine and Dexamethasone as a preventative measure. 



The gases used in advanced ultrasound contrast agents can be of concern under some 
conditions, due to their relatively slow dissolution. In rodents, injection of gas-carrier 
contrast agents can induce intestinal and hepatic lesions due to the growth of relatively 
large gas bubbles (Rasmussen et al., 2003; Dirven et al., 2003). In addition, intra-arterial 
injection of a contrast agent can lead to local blood-brain barrier disruption, without any 
ultrasound exposure (Mychaskiw et al., 2000). The mechanism of this effect is uncertain, 
but may be related to relatively large gas bodies containing perflutren gas, which can enter 
the brain directly by intra-arterial injection without the filtering effect of lung passage. 



Intra-arterial 
injection of 
contrast agents is 
not recommended 
for CEDUS 
examinations 



Optison and Definity were approved for patients with suboptimal echocardiograms in 
1997 and 2001, respectively. In post-marketing surveillance, serious adverse events were 
reported, prompting a re-evaluation of the labelling for these agents in 2007. A "black box" 
warning of serious cardiopulmonary reactions and contraindications were added to the 
package inserts. The contraindications were later relaxed, but warnings remain for serious 
cardiopulmonary reactions, anaphylactoid reactions, systemic embolization in patients 
with cardiac shunts, high ultrasound mechanical index (MI) (greater than 0.8) and QTc 
prolongation in the electrocardiogram (ECG). These are noteworthy warnings for clinical 
diagnostic ultrasound. For example, QTc prolongation has recently become an important 
factor for drug safety evaluation (Shah, 2005; Whellan et al., 2009). The guideline used 
by the FDA Center for Drug Evaluation and Research is that a QTc prolongation >20ms 
may be of concern (CDER, 2005). The package insert for Definity lists a warning that in 
a preclinical study, a QTc prolongation >30ms was noted in 29% of subjects (64/221). 
The revised package inserts are available on line at FDA (2008a) for Optison and FDA 
(2008b) for Definity. Sonovue was approved by the European Medicines Agency in 2001. 
A post-marketing review in 2004 resulted in an Urgent Safety Restriction issued in 2004, 
including suppression of the echocardiography indication. A scientific discussion is 
available on line at EMA (2004). The restrictions were later relaxed, and a new European 
Public Assessment Report issued (EMA, 2008). 



Advanced 
ultrasound 
contrast agents 
have been 
the subject 
of regulatory 
review for serious 
adverse events 
reported in 
post-marketing 
surveillance 



8.3.2 Recent epidemiological studies 

The restrictions placed in the labelling of contrast agents have been controversial, 
and have stimulated several retrospective epidemiological studies of possible adverse 
effects. A study was conducted using records of 23,188 investigations using Sonovue 

109 



8 The safe use of contrast-enhanced diagnostic ultrasound 



Recent 
epidemiological 
studies are 
reassuring that 
the 24 h mortality 
rates were similar 
for patients 
receiving contrast 
or non-contrast 
ultrasound 
examinations 



in abdominal applications (Piscaglia et al., 2006). Adverse events were reported in 
only 29 cases. No information was available on the prevalence of cardiac disease, the 
reasons for the examinations, the dose or ultrasound imaging utilized. No comparison 
was made with non-contrast patients. Kusnetzky et al. (2008) investigated the incidence 
of death within 24h of echocardiography examination in records of 12,475 patients 
without contrast and 6,196 with Definity. The death rate was 0.37% without contrast 
and 0.42% with contrast, which was not a statistically significant increase. Little patient 
information, such as the reason for the examinations, was available. A large database 
was used to examine 24h mortality after echocardiography in a study supported by 
Lantheus Medical Imaging (Main et al., 2008). Echocardiography was performed during 
hospitalization in 4,242,712 patients without contrast enhancement and 58,254 (1.4% of 
the total) with Definity enhancement. The contrast dosage or ultrasound examination 
methods were not specified. The 24h mortality rates were 1.08% and 1.06% for non- 
contrast and contrast echocardiography patients, respectively. Multivariate logistic 
regression analysis indicated that contrast echocardiography patients were 24% less 
likely to die than patients not receiving a contrast agent (adjusted odds ratio 0.76, 95% 
confidence interval 0.70-0.82). However, only crude mortality data (not cause of death) 
were available for the patients, and the vast majority of patients did not receive ultrasound 
contrast agent due to clinical contraindications. Records of 5,250 consecutive adult 
patients were evaluated prospectively for adverse events after dobutamine-atropine 
stress echocardiography and myocardial contrast echocardiography (MCE) with Sonovue 
(Aggeli et al., 2008). Patients with unstable angina, acute coronary syndrome in the 
previous 30 days and other similar criteria were excluded. An average of 2.5 ml of Sonovue 
was given within lmin, and 1.7MHz real-time MCE [mechanical index (MI)=0.1-0.2] 
was used with intermittent agent-destructive scans (MI=1.7). There were no reported 
deaths. A variety of adverse events was reported, with the most prevalent event being 
dry-mouth (19.8%). The total of cardiac arrhythmia events was 6.3%, including 4% 
incidence of premature ventricular contractions. There was no comparison with non- 
contrast patient groups. 



8.4 Research on biological effects induced by CEDUS 



8.4.1 Bio-effects in vitro 



The thresholds 
for membrane 
injury and lysis in 
monolayer cells 
have been shown 
to be very low 
and proportional 
to diagnostic 
ultrasonic 
frequency 



Many in vitro studies of cavitation bio-effects have taken advantage of the nucleation 
ability of ultrasound contrast agents to improve and enhance results (Miller, 2007). 
These studies reveal the range of potential cellular bio-effects with contrast agents, but 
results are not necessarily directly applicable to in vivo conditions. Only a few studies 
have employed actual clinical ultrasound machines. Miller and Quddus (2000a) studied 
cell membrane bio-effects in epidermoid cell monolayers scanned using Optison and 
3.5 MHz diagnostic ultrasound. The monolayer was located at the top of the vessel 
during exposure so that the gas bodies would rise to become adjacent to the monolayer. 
Sonoporation (transient membrane permeabilization) was detected at RPAs as low as 
0.23 MPa in pulsed Doppler mode and 0.39MPa in B-mode. Phagocytic cells grown in 
monolayers and pre-incubated with Optison to attach the gas bodies to the cells were 
lysed by exposure to ultrasound produced by a diagnostic ultrasound machine operated 

110 



The safe use of contrast-enhanced diagnostic ultrasound 8 



in spectral Doppler mode, with an RPA threshold of ~0.2MPa (Miller and Quddus 2001). 
In similar experiments using 2-cycle ultrasound pulses, the FiPA thresholds for killing 
phagocytic cells pre-loaded with Optison showed a linear correlation (r 2 =0.982) with 
frequency over the l-10MHz range, increasing with a slope of -0.06 MPa/MHz (Miller 
and Dou 2004a). These data are plotte d in Figure 8.1. These pressure thresholds were lower 
than those for nucleation of inertial cavitation and were proportional to the frequency 
(not its square root, as expected from the MI, see Figure 8.1). The frequency dependence 
of gas-body destabilization and cellular bio-effects observed for this in vitro system can 
be modelled by theory for shell stresses and acoustic microstreaming shear stress on 
cells (Miller and Dou, 2004b). The theory substantiated the observed linear dependence 
of thresholds on frequency. Owing to the design of this model monolayer system for 
maximum sensitivity to cellular bio-effects, the thresholds observed may approximate the 
lowest RPAs for which biologically significant bio-effects (i.e. cell killing) can be expected 
for CEDUS. 




12 5 10 

Frequency (MHz) 



Figure 8.1. The frequency dependence of in vitro lysis of monolayer cells (diamonds) with 
contrast agent gas bodies in contact with the cells, determined using simulated diagnostic 
exposure. Thresholds for glomerular capillary haemorrhage in rat kidney are also plotted 
for diagnostic ultrasound exposure (open circles) and simulated diagnostic ultrasound (filled 
circles). The straight lines were fitted to the data by the least-squares method and indicate 
a strong dependence of CEDUS bio-effects thresholds (approximate proportionality to 
frequency). For comparison, the US-FDA maximum value of the on-screen mechanical index 
(Ml = 1 .9) for diagnostic scanners is also plotted. 

111 



8 The safe use of contrast-enhanced diagnostic ultrasound 



8.4.2 Early research on CEDUS induced bio-effects in vivo 



Capillary leakage 
and rupture in 
intestine and 
skeletal muscle 
has been shown 
to occur in mice 
for CEDUS, even 
for single image 
frames 



For pulsed ultrasound, the presence of Albunex gas bodies in the circulation enhanced 
the induction of petechiae in mouse intestine (Miller and Gies, 1998). The occurrence of 
microvascular injury to rat mesentery was examined by Kobyashi et al. (2002, 2003) using 
diagnostic ultrasound in an intra-vital preparation. A phased array probe was used at 
1.8MHz RPAs of 0.14MPa (equivalent mechanical index, eMI -0.10) or 0.82MPa (eMI 
-0.61) with Levovist or Definity. Capillary ruptures and endothelial cell killing were 
observed in capillaries and venules (but not arterioles) for real-time imaging even for the 
low-MI setting. Micro-vessel rupture was observed to occur with intra-vital microscopy 
in rat spinotrapezius muscle (Skyba et al., 1998). Optison gas bodies were injected and 
exposed with a 2.3 MHz diagnostic ultrasound scanner in the harmonic imaging mode 
resulted in capillary rupture and non-viable cells for eMI values above 0.4. The induction 
of petechiae and capillary leakage by CEDUS was found in skeletal muscle in vivo by 
Miller and Quddus (2000b). The beam from a 2.5MHz diagnostic probe was aimed at the 
abdomens of anesthetized mice mounted in a water bath to provide conditions simulating 
human exposure. Muscle petechiae were significantly elevated relative to shams at RPAs 
above 0.64MPa (eMI=0.4). A single image frame was sufficient to produce petechiae 
and capillary rupture was also seen in fat, small intestine and Peyers' patches (intestinal 
lymph nodes). 



8.4.3 Bio-effects in contrast-enhanced echocardiography 



MCE with 
intermittent 
contrast 
destruction can 
induce ECG 
premature 
complexes in 
animals and 
humans 



Ultrasound contrast agents were initially applied in echocardiography to opacify the left 
ventricular chamber and improve endocardial border delineation, which remains, at this 
time, the only indication approved in the USA. In addition, MCE can be used to image 
perfusion, which includes intermittent scans to destroy gas bodies in the myocardial 
microcirculation (Porter and Xie, 2010). During MCE, increased numbers of premature 
complexes (PCs) in the ECG were reported in humans with an experimental contrast 
agent by van der Wouw et al. (2000). Intermittent imaging was conducted at 1.66 MHz 
with a significant increase in PCs to about 1 per minute seen for end-systolic triggering 
at MI = 1.5, but not at MI = 1.1. Subsequently, this PC-effect of MCE has been reported 
independently (confirming the observation) in humans using experimental contrast 
agents (Chapman et al., 2005), in dogs using a commercial agent (Okazaki et al., 2004; Miller 
et al, 2006), and also in rats with various agents (e.g. Li et al, 2003, 2004; Vancraeynest 
et al, 2009). The PCs from MCE are important because these provide a clinically observable 
indication of myocardial perturbation. 



MCE can induce 
microvascular 
bio-effects 
including capillary 
leakage, capillary 
haemorrhage, 
release of cardiac 
injury biomarkers, 
and transient 
arrhythmia 



The microvascular effects of MCE have been examined in several different animal models. 
Chen et al. (2002) imaged rat hearts at 1.3 MHz with ECG triggering each 4 cardiac cycles 
with Optison or Definity, and found elevations of Troponin T in blood plasma, indicating 
myocardial damage, after 30min for Mis of 1.2 and 1.6. Clinical studies of MCE were 
conducted with Optison and power Doppler imaging at 1.7-1.9 MHz each 1-3 heartbeats at 
end systole with no consistent changes in cardiac injury markers (Borges et al., 2002; Knebel 
et al., 2005). However, Vancraeynest et al. (2007) detected a signifi <ant release of cardiac 



112 



The safe use of contrast-enhanced diagnostic ultrasound 8 




Figure 8.2. Ultrasound images for an open-chest canine model of MCE with intermittent 
imaging each 4 beats at end systole. The image on the left shows the normal B-mode image. The 
image on the right shows the same view after contrast-agent infusion and the accompanying 
ECG trace shows two PCs occurring just after the image triggers (vertical spikes). 

injury biomarkers in humans, including troponin I in a group with 1.3 MHz MCE at an MI 
of 1.5. In rats, MCE at 1.7MHz with Optison, Definity or Imagent was reported to induce 
microvascular permeabilization and PCs (Li et al., 2003, 2004). Threshold RPAs above which 
effects were significant were about the same for all three agents, and were l.OMPa (eMI=0.8) 
for PCs and 0.54 MPa (eMI = 0.41) for capillary haemorrhage. Similar bio-effects were seen 
in a canine model of MCE (Miller et al., 2006), as shown for open-chest imaging in Figure 
8.2. The resulting leakage of blue dye and petechiae is shown in Figure 8.3a. Vancraeynest 
et al. (2009) reported that triggered imaging at high MI with multiple doses of contrast agent 
over periods up to 30 min resulted in left ventricular dysfunction, ST-segment elevation and 
even the death of some rats (up to 5 of 6 rats at the longest duration). 



Histologically defined microlesions with inflammatory cell infiltration induced by MCE 
were reported by Miller et al. (2005a). MCE with 1:4 end-systolic triggering was performed 
in rats at 1.5 MHz and 2 MPa (eMI = 1.7) with Optison, which induced microvascular 
leakage, petechiae and cardiomyocytes with contraction band necrosis. Similar bio- 
effects were found histologically by Vancraeynest et al. (2006). The lethal cardiomyocyte 
injury could be characterized through use of Evans blue dye as a vital stain, as shown 
in Figure 8.3b (Miller et al., 2005b). The thresholds and exposure response trends were 
found to be quite similar for PCs and cardiomyocyte death, as shown in Figure 8.4, with a 
good correlation between these bio-effects over a wide range of experimental conditions 
(Miller et al., 2011). This result was consistent with the hypothesis that the PCs were a 
physiological response to the lethal cardiomyocyte injury. 



PCs induced by MCE 
were associated 
with lethal 
cardiomyocyte injury 
and microlesion 
formation within the 
myocardium 



8.4.4 Bio-effects of CEDUS in kidney 

The kidney appears to be an organ which is especially sensitive to CEDUS associated 
bio-effects, owing to the unique structure of the glomerular capillaries which filter 
liquid into the urinary space and tubules. Wible et al. (2002) found that glomerular 



113 



8 The safe use of contrast-enhanced diagnostic ultrasound 





Figure 8.3. A canine heart (left) after exposure in an open-chest canine model of MCE showed 
petechial haemorrhages and leakage of Evans blue dye in the scan plane (scale bar 5 mm). MCE 
at relatively high RPAs induced lethal cardiomyocyte injury (right) indicated by the fluorescent 
red staining in a rat heart after 1 day. The nuclei of all cells present are indicated by the 
fluorescent blue staining. 



800 



| 600 

o 
O 

0 
O 

H — ' 

£Z 
0 
O 
c/) 

CD 
i— 

O 

I 200 



400 



-#— Fluorescent Cell Count 
■Q— Premature Complexes 



01 

0.0 




80 



60 



C/5 

0 

X 

0 



CL 

E 
o 
O 

40 0 

3 

E 
0 

20 £ 



0.5 1.0 1.5 

Mechanical Index (On Screen Equivalent) 



1.9 



Figure 8.4. The fluorescent-cell counts and PCs found for MCE of rat hearts, plotted against 
the eMI. 



GCH induced 
by CEDUS 
can obstruct 
and injure the 
tubules, and 
produce clinically 
detectable 
haematuria 



capillary haemorrhage (GCH) was induced by CEDUS in rats. Ultrasound contrast 
agents including Optison were used with intermittent imaging. GCH into the proximal 
convoluted tubules was visible on the kidney surface within the scanned plane, and was 
statistically significant for an RPA of 1.26 (eMI = 0.94) at 1.8 MHz. A 1.5 MHz diagnostic 
exposure system was designed to simulate human clinical exposure in rats (Miller et ah, 
2007a) . Exposure at 1 .8 MPa with a 1 s trigger interval for 1 min during infusion of Definity 

114 



The safe use of contrast-enhanced diagnostic ultrasound 8 




Figure 8.5. (a) A rat kidney after contrast-enhanced diagnostic ultrasound showing the 
pattern of petechia within the scan plane, which were visible on the surface of the kidney, 
(b) A histological section with a Bowman's space and proximal tubule (arrows) positive for 
glomerular capillary haemorrhage (GH) and a normal glomerulus (NG) (scale bar 0.1 mm). 



produced of 37% ± 5.0% of the glomeruli in histology at the scan plane, which decreased 
to an apparent threshold of 0.73 MPa (eMI = 0.6). Examples of the visible haemorrhage 
on the kidney surface, and of the histological appearance of the GCH are shown in Figure 
8.5. The agent dosage used for this study in rats followed the human recommendation in 
the package insert, but the circulating gas-body dose may have been much less than that 
which would be delivered in humans, owing to substantial gas-body loss in the small 
animals (Miller et al., 2010b). The kidney CEDUS in rats produced readily detectable 
haematuria, which parallelled the incidence of GCH as shown in Figure 8.6 (Williams 
et al., 2007). In addition, many of the injured nephrons remained filled with tightly 
packed erythrocytes 24 h after imaging, with the degeneration seen in acute tubular 
necrosis. Histological observation of Bowman's space showed enlargement and clots, 
which were indicative of tubular obstruction (Miller et al., 2009). For kidney CEDUS in 
swine, Jimenez et al. (2008) reported that GCH did not occur with Sonovue; however, 
the ultrasound probe was placed directly on the kidney, which located the cortex in the 
near-field of the probe with very low RPAs. In contrast, a study in swine showed that 
CEDUS using Definity did produce GCH in the focal zone, which was comparable to 
that seen in rats (Miller et al., 2010a). 



The frequency dependence of thresholds for GCH induced by CEDUS was examined in 
rats (Miller et al., 2008b). Diagnostic ultrasound scanners were used for exposure at 1.5, 2.5, 
3.2, 5.0 and 7.4 MHz and a laboratory exposure system was used at 1.0, 1.5, 2.25, 3.5, 5.0 
and 7.5 MHz. The RPA thresholds for GCH were proportional to the ultrasound frequency 
(not its square root) at 0.5 MPa/MHz for diagnostic ultrasound and 0.6 MPa/MHz for 
the laboratory system, as shown in Figure 8.1. These results show that the frequency 
dependence of the on-screen MI does not have the correct frequency dependence for 
anticipating the GCH bio-effect. 



Thresholds 
for GCH have 
a different 
frequency 
dependence than 
the Ml 



The influence of imaging mode on GCH was investigated in rat kidney (Miller et al., 
2007b). B-mode flash echo imaging (FEI), colour Doppler (CD) FEI and real-time Doppler 
imaging at 1 frame per second were compared for 2.5 MHz and 2.6 MPa with Definity. 
B-mode induced 38.6% ±17.1% GCH, while the CD mode gave 19.6% ±7.4% GCH 

115 



8 The safe use of contrast-enhanced diagnostic ultrasound 



CD 
CD 



CD 
CD 

X 

_CD 

i_ 

CD 

E 
o 

CD 



50 



40 



-#— Haemorrhage 
■Q— Haematuria 




1.5 

Mechanical Index (On Screen Equivalent) 



1.9 



Figure 8.6. The percentage of glomeruli in the scan plane with evidence of capillary 
haemorrhage and the relative haematuria score after CEDUS in rats, plotted against the eMI. 



The GCH 
bio-effect can 
be reduced by 
using special 
pulse sequences 
for contrast 
destruction 



(P<0.02) and the Doppler mode gave 5.3% ±3.8% GCH (P< 0.001), respectively. This 
result was surprising, because the Doppler mode delivered pulses for about 83.5 ms 
per image, compared to 15.8 ms and 0.53 ms for CD and B-modes, respectively. This 
finding of reduced GCH for slow frame rates suggests that GCH could be minimized in 
CEDUS examinations by using specially designed pulse-amplitude sequences for agent 
destruction scans. 



CEDUS did 
not increase 
metastasis in 
subcutaneous 
mouse tumours 



CEDUS may 
induce platelet, 
hepatic or 
endothelial cell 
perturbation in 
liver 



8.4.5 Bio-effects of CEDUS in tumours, liver and brain 

Contrast-enhanced ultrasound scanning of various tissues can assist in the identification 
of malignant tumours but might also cause microvascular perturbations. Subcutaneous 
melanoma tumours in mice, which have an enhanced potential for lung metastasis, were 
scanned intermittently with 1.5 MHz diagnostic ultrasound during or after Definity 
injection (Miller and Dou, 2005). For ultrasound plus contrast agent, observation of 
a brightening of the tumour image confirmed the interaction of ultrasound with the 
contrast agent within the tumour. One day after scanning, the primary tumour was 
surgically removed, and the possible lung metastasis allowed to develop for 28 days. No 
significant increase in metastases was seen in the lungs. 

The liver is often the subject of ultrasound examinations, and these can be improved by 
CEDUS. Effects of Levovist-aided ultrasound on rat liver was investigated by Shigeta et al. 
(2004). The on-screen MI values were 1.8 at 8 MHz and 0.7 at 12 MHz, and both were used 

116 



The safe use of contrast-enhanced diagnostic ultrasound 8 



on each rat. The transducers were moved to expose the entire liver, which was examined 
by electron microscopy. Qualitative observation of the specimens revealed increased 
platelet aggregation and endothelial cell damage in the sinusoids for ultrasound plus 
contrast agent groups. Levovist was compared to an experimental agent by Shigeta et al. 
(2005). The gas bodies appeared to be taken up by some Kupff er (phagocytic liver cells) 
cells. In addition, the hepatic cells had distinct vascularization not seen in the sham or 
contrast only groups. 



The possible alteration of the blood-brain barrier by contrast aided ultrasound was 
investigated by Schlachetzki et al. (2002) using Levovist and Optison and by using 
Sonovue. Transcranial colour coded sonography was performed on human volunteers 
using a 2-3.5 MHz phased array probe with maximal output settings. The contrast 
agent Magnevist was also injected intravenously, and evidence of microvascular leakage 
was sought using magnetic resonance imaging of the brain. There were no indications 
of focal signal enhancement attributable to extravasation of the Magnevist. Similar 
clinical research was conducted using Sonovue in patients with small vessel disease 
using intermittent 2.5 MHz imaging with a mean MI of 0.7, and no blood-brain barrier 
changes were detectable with contrast magnetic resonance imaging (Jungehulsing 
et al., 2008). 



Transcranial 
CEDUS does not 
appear to induce 
microvascular 
bio-effects in 
brain 



CEDUS induced bio-effects may present opportunities for therapeutic benefits under 
special circumstances. Pulsed Doppler ultrasound is used in diagnosis of stroke 
patients. In patients with occluded middle cerebral artery being treated with tPA, 
continuous transcranial Doppler monitoring has been shown to improve the outcome 
of the treatment relative to placebo monitoring (Alexandrov et al., 2004). This suggests 
that the ultrasound accelerated the thrombolysis. Ultrasound contrast agents may be 
used to improve transcranial signals, and addition of contrast agent has been tested 
for the continuous monitoring treatment with encouraging results (Molina et al., 2006; 
Perren et al., 2008). For example, 2 h of 2 MHz continuous Doppler monitoring during 
tPA infusion was augmented by Levovist injection in three doses at 2, 20 and 40 min. 
The 2 h complete recanalization rate was statistically significantly higher in the CEDUS 
group (54.5%) relative to non-contrast Doppler (40.8%) and to tPA alone (23.9%), with no 
trend in observed intracranial haemorrhage. The future development of these intriguing 
stroke treatment methods are presently uncertain, owing to many possible variations in 
ultrasound application, contrast agent, thrombolytic drug and side effects (Rubiera and 
Alexandrov, 2010). 



CEDUS-enhanced 
thrombolysis may 
find beneficial 
application in 
stroke treatment 



8.5 Discussion 



Ultrasound contrast agents are suspensions of gas bodies prepared as an injectable 
drug. The gas bodies are engineered with stabilizing shells or lipid skins and contain 
slowly dissolving gasses such as perfluoropropane or sulphur hexafluoride. This design 
provides useful persistence times for the micron-sized gas bodies after intravenous 
injection, and strong echoes for imaging. CEDUS is valuable for echocardiography and 
abdominal imaging particularly of the liver and kidneys. CEDUS is primarily used to 
improve diagnostic value after sub optimal examinations. In addition, these agents also 

117 



Contrast 
enhancement 
brings 
completely 
new imaging 
capabilities 
to diagnostic 
ultrasound 



8 The safe use of contrast-enhanced diagnostic ultrasound 



bring a new dimension to diagnostic ultrasound by allowing the imaging of blood flow 
and tissue perfusion without ionizing radiation. 



Ultrasound 
contrast agents 
can induce 
adverse patient 
reactions and 
cavitation 
activity 
generated 
by diagnostic 
ultrasound 
pulses can 
induce a range 
of harmful 
microscale 
bio-effects 



CEDUS also brings completely new patient risks to diagnostic ultrasound. In post- 
marketing surveillance, contrast agents have been associated with adverse patient 
reactions. These findings have lead to new contraindications and warning in package 
inserts. Worrisome concerns include cardiac arrhythmia and anaphylactoid reactions. 
Recent epidemiology studies have been reassuring in regard to possible fatal reactions 
within one day, but no information is available from large randomized controlled 
trials. In addition, destabilization of the gas bodies effectively nucleates ultrasonic 
cavitation from liberated microbubbles, which provides a potent mechanism for 
biological effects not normally associated with diagnostic ultrasound. For low pulse- 
pressure amplitudes, cellular injury and capillary leakage may occur. With intermittent 
imaging at high diagnostic pressure amplitudes, lethal injury of cardiomyocytes can 
occur with accompanying PCs in the ECG. In kidney, glomerular capillary rupture can 
occur leading to tubular obstruction and haematuria. Further research into the possible 
medical significance of these microscale bio-effects would be valuable for patient risk 
assessment. 



The on-screen 
Ml provides 
little dosimetric 
guidance for 
the safe use of 
CEDUS aside 
from a general 
indication 
of exposure 
within a 
specific 
examination 



The dosimetric characterization of CEDUS examinations has proven to be complex. The 
theoretical MI and its upper limit for diagnostic ultrasound were established without 
regard for any bio-effects. This exposure index does not relate to the destabilization of the 
microbubbles (Forsberg et al., 2005) or the frequency dependence of microlesions, which 
occur well below the upper limit for diagnostic ultrasound, see Figure 8.1 (Miller et al., 
2008a,b). Bio-effects depend on FiPA, agent dose, agent delivery, image mode and tissue 
properties. Low-MI imaging modes can lead to tissue injury through the use of high 
RPAs in agent clearance scans. Further consideration of CEDUS by regulatory agencies 
to develop better dosimetric parameters, to recommend specific agent-destructive scans 
that minimize bio-effects risk and to provide specific guidance on microscale bio-effects 
all would be of value for the advancement of CEDUS. 



With training 
and attention 
to safety issues, 
the benefits of 
CEDUS can be 
delivered with 
a minimum of 
patient risk 



8.6 Conclusion 

Physicians and sonographers should possess a general knowledge of CEDUS safety 
issues in order to maximize the risk-to-benefit ratio in medical imaging. Specifically, 
safety issues should be covered in training or certification for CEDUS. Guidance for 
the performance CEDUS is available in recent publications from the American Society 
of Echocardiography (Mulvagh et al., 2008) and the European Federation of Societies 
for Ultrasound in Medicine and Biology (Claudon et al., 2008). Strategies for reducing 
possible ultrasound-induced bio-effects may be found in recent safety reviews (Barnett 
et al, 2007; Miller et al, 2008a; ter Haar, 2009). For a specific CEDUS examination the 
bio-effects risk may be minimized generally by keeping the MI for both imaging and 
contrast destruction less than 0.4, above which level bio-effects have been observed 
(AIUM, 2008). Finally, the contraindications, warnings and usage instructions found 
in package inserts of ultrasound contrast agents should be followed diligently to 
minimize patient risks. 

118 



The safe use of contrast-enhanced diagnostic ultrasound 8 



References 

AIUM. 2008. Bioeffects of Diagnostic Ultrasound with Gas Body Contrast Agents. Laurel, 
MD: American Institute of Ultrasound in Medicine. Available online at: http://www.aium. 
org/publications/statements.aspx. 

Aggeli C, Giannopoulos G, Roussakis G, Christoforatou E, Marinos G, Toli C, et al. 
2008. Safety of myocardial flash-contrast echocardiography in combination with 
dobutamine stress testing for the detection of ischaemia in 5250 studies. Heart, 94, 
1571-1577. 

Alexandrov AV, Molina CA, Grotta JC, Garami Z, Ford SR, Alvarez-Sabin J, et al. 2004. 
Ultrasound-enhanced systemic thrombolysis for acute ischemic stroke. N Engl ] Med, 351, 
2170-2178. 

Apfel RE. 1982. Acoustic cavitation: a possible consequence of biomedical uses of 
ultrasound. Br } Cancer Suppl, 45, 140-146. 

Apfel RE, Holland CK. 1991. Gauging the likelihood of cavitation from short-pulse, low- 
duty cycle diagnostic ultrasound. Ultrasound Med Biol, 17, 179-185. 

Averkiou M, Powers J, Skyba D, Bruce M, Jensen S. 2003. Ultrasound contrast imaging 
research. Ultrasound Q, 19, 27-37. 

Barnett SB, Duck F, Ziskin M, 2007. Recommendations on the safe use of ultrasound 
contrast agents. Ultrasound Med Biol, 33, 173-174. 

Bleeker HJ, Shung KK, Barnhart JL. 1990. Ultrasonic characterization of Albunex, a new 
contrast agent. JAcoust Soc Am, 87, 1792-1797. 

Borges AC, Walde T, Reibis RK, Grohmann A, Ziebig R, Rutsch W, et al. 2002. Does 
contrast echocardiography with Optison induce myocardial necrosis in humans? / Am 
Soc Echocardiogr, 10, 1080-1086. 

Bramos D, Tsirikos N, Kottis G, Pamboucas C, Kostopoulou V, Trika C. 2008. The acute 
effect of an echo-contrast agent on right ventricular dimensions and contractility in pigs. 
/ Cardiovasc Pharmacol. 51, 86-91. 

Carstensen EL, Gracewski S, Dalecki D. 2000. The search for cavitation in vivo. Ultrasound 
Med Biol, 26, 1377-1385. 

CDER. 2005. Guidance for Industry. E14 Clinical Evaluation of QT/QTc Interval Prolon- 
gation and Proarrhythmic Potential for Non-antiarrhythmic Drugs. Center for Drug 
Evaluation and Research. Rockville, MD: Food and Drug Administration. Available at: 
http://www.fda.gov/downloads/RegulatoryInformation/Guidances/UCM129357.pdf. 
Chapman S, Windle J, Xie F, McGrain A, Porter TR. 2005. Incidence of cardiac arrhythmias 
with therapeutic versus diagnostic ultrasound and intravenous microbubbles. / Ultrasound 
Med, 24, 1099-1107. 

Chen S, Kroll MH, Shohet RV, Frenkel P, Mayer SA, Grayburn PA. 2002. Bioeffects of 
myocardial contrast microbubble destruction by echocardiography. Echocardiography, 19, 
495-500. 

Chomas JE, Dayton P, May D, Ferrara K. 2001. Threshold of fragmentation for ultrasonic 
contrast agents. / Biomed Optics, 6, 141-150. 

Christiansen C, Kryvi H, Sontum PC, Skotland T. 1994. Physical and biochemical 
characterization of Albunex™, a new ultrasound contrast agent consisting of air-filled 
albumin microspheres suspended in a solution of human albumin. Biotechnol Appl 
Biochem, 19, 307-320. 

119 



8 The safe use of contrast-enhanced diagnostic ultrasound 



Church CC. 2002. Spontaneous homogeneous nucleation, inertial cavitation and the 
safety of diagnostic ultrasound. Ultrasound Med Biol, 28, 1349-1364. 

Claudon M, Cosgrove D, Albrecht T, Bolondi L, Bosio M, Calliada F, et al. 2008. Guidelines 
and good clinical practice recommendations for contrast enhanced ultrasound (CEUS) - 
update 2008. Ultraschall Med, 29, 28^4. 

Cosgrove D, Blomley M. 2004. Liver tumors: evaluation with contrast-enhanced 
ultrasound. Abdom Imaging, 29, 446-454. 

Cosgrove D. 2006. Ultrasound contrast agents: an overview. Eur ] Radiol, 60, 324-330. 
Dirven HA, Rasmussen H, Johnsen H, Videm S, Walday P, Grant D. 2003. Intestinal and 
hepatic lesions in mice, rats, and other laboratory animals after intravenous administration 
of gas-carrier contrast agents used in ultrasound imaging. Toxicol Appl Pharmacol, 188, 
165-175. 

EMA. 2004. Sonovue Scientific Discussion. European Medicines Agency, EU. Available 
at: http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Scientific_ 
Discussion/human/000303/WC500055376.pdf. 

EMA. 2008. Sonovue Summary & Label. European Medicines Agency, EU. Available 
at: http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_ 
Information/human/000303/WC500055380.pdf. 

FDA, 2008a. Optison. Food and Drug Administration, USA. Available at: http://www. 
accessdata.fda.gov/drugsatfda_docs/label/2008/020899s0111bl.pdf. 

FDA. 2008b. Defirity. Food and Drug Administration, USA. Available at: http://www. 
accessdata.fda.gov/drugsatfda_docs/label/2008/021064s0091bl.pdf. 

Flynn HG. 1982. Generation of transient cavitation in liquids by microsecond pulses of 
ultrasound. J Acoust SocAm, 72, 1926-1932. 

Forsberg F, Shi WT, Goldberg BB. 2000. Subharmonic imaging of contrast agents. 
Ultrasonics, 38, 93-98. 

Forsberg F, Shi WT, Merritt CR, Dai Q, Solcova M, Goldberg BB. 2005. On the usefulness 
of the mechanical index displayed on clinical ultrasound scanners for predicting contrast 
microbubble destruction. / Ultrasound Med, 24, 443-450. 

Goldberg BB, Raichlen JS, Forsberg F (editors). 2001. Ultrasound Contrast Agents Basic 
Principles and Clinical Applications. 2nd Edition. London, UK: Martin Dunitz, pp. 1-416. 
Gramiak R. 1997. The beginnings of ultrasound contrast. In Ultrasound Contrast Agents, 
Goldberg BB (editor). London, UK: Martin Dunitz, pp. 1-8. 

Grauer SE, Pantely GA, Xu J, Giraud GD, Shiota T, Sahn DJ. 1996. Myocardial imaging 
with a new transpulmonary lipid-fluorocarbon echo contrast agent: experimental studies 
in pigs. Am Heart }, 132, 938-945. 

ter Haar G. 2009. Safety and bio-effects of ultrasound contrast agents. Med Biol Eng 
Comput, 47, 893-900. 

Jimenez C, de Gracia R, Aguilera A, Alonso S, Cirugeda A, Benito J, et al. 2008. In situ 
kidney insonation with microbubble contrast agents does not cause renal tissue damage 
in a porcine model. / Ultrasound Med, 27, 1607-1615. 

Jungehulsing GJ, Brunecker P, Nolte CH, Fiebach JB, Kunze C, Doepp F, et al. 2008. 
Diagnostic transcranial ultrasound perfusion-imaging at 2.5 MHz does not affect the 
blood-brain barrier. Ultrasound Med Biol, 34, 147-150. 

Killam A, Dittrich HC. 1997. Cardiac applications of Albunex and FS069. In Ultrasound 
Contrast Agents, Goldberg BB (editor). London, UK: Martin Dunitz Ltd, pp. 43-55. 

120 



The safe use of contrast-enhanced diagnostic ultrasound 8 



Knebel F, Schimke I, Eddicks S, Walde T, Ziebig R, Schattke S, et al. 2005. Does contrast 
echocardiography induce increases in markers of myocardial necrosis, inflammation and 
oxidative stress suggesting myocardial injury? Cardiovasc Ultrasound, 3, 21. 
Kobayashi N, Yasu T, Yamada S, Kudo N, Kuroki M, Kawakami M, et al. 2002. Endothelial 
cell injury in venule and capillary induced by contrast ultrasonography. Ultrasound Med 
Biol, 28, 949-956. 

Kobayashi N, Yasu T, Yamada S, Kudo N, Kuroki M, Miyatake K, et al. 2003. Influence of 
contrast ultrasonography with perflutren lipid microspheres on microvessel injury. Circ 
J, 67, 630-636. 

Krishna PD, Newhouse VL. 1997. Second harmonic characteristics of the ultrasound 

contrast agents Albunex and FS069. Ultrasound Med Biol, 23, 453-459. 

Kusnetzky LL, Khalid A, Khumri TM, Moe TG, Jones PG, Main ML. 2008. Acute 

mortality in hospitalized patients undergoing echocardiography with and without an 

ultrasound contrast agent: results in 18,671 consecutive studies. ] Am Coll Cardiol, 51, 

1704-1706. 

Li P, Cao LQ, Dou CY, Armstrong WR, Miller DL. 2003. Impact of myocardial contrast 
echocardiography on vascular permeability: an in vivo dose response study of delivery 
mode, ultrasound power and contrast dose. Ultrasound Med Biol, 29, 1341-1349. 
Li P, Armstrong WR, Miller DL. 2004. Impact of myocardial contrast echocardiography 
on vascular permeability: comparison of three different contrast agents. Ultrasound Med 
Biol, 30, 83-91. 

Main ML, Ryan AC, Davis TE, Albano MP, Kusnetzky LL, Hibberd M. 2008. Acute 
mortality in hospitalized patients undergoing echocardiography with and without an 
ultrasound contrast agent (multicenter registry results in 4,300,966 consecutive patients). 
Am ] Cardiol, 102, 1742-1746. 

Marmottant P, van der Meer S, Emmer M, Versluis M, de Jong N, Hilgenfeldt S, et al. 2005. 
A model for large amplitude oscillations of coated bubbles accounting for buckling and 
rupture. / Acoust Soc Am, 118, 3499-3505. 

Mattrey RF, Pelura TJ. 1997. Perfluorocarbon-based ultrasound contrast agents. In 
Ultrasound Contrast Agents, Goldberg BB (editor). London, UK: Martin Dunitz Ltd, 
pp. 83-99. 

Miller DL. 2000. The safety of ultrasound contrast agents. In The Safe Use of Ultrasound 
in Medical Diagnosis, ter Haar G, Duck FA (editors). 2nd Edition. London, UK: British 
Medical Ultrasound Society & The British Institute of Radiology, pp. 72-85. 
Miller DL. 2007. WFUMB safety symposium on echo-contrast agents: in vitro bioeffects. 
Ultrasound Med Biol, 33, 197-204. 

Miller DL, Dou C. 2004a. Membrane damage thresholds for 1- to 10-MHz pulsed 
ultrasound exposure of phagocytic cells loaded with contrast agent gas bodies in vitro. 
Ultrasound Med Biol, 30, 973-977. 

Miller DL, Dou C. 2004b. Theoretical gas body pulsation in relation to empirical gas- 
body destabilization and to cell membrane damage thresholds. / Acoust Soc Am, 116, 
3742-3749. 

Miller DL, Dou C. 2005. Contrast-aided diagnostic ultrasound does not enhance lung 
metastasis in a mouse melanoma tumor model. / Ultrasound Med, 24, 349-354. 
Miller DL, Gies RA. 1998. Gas-body-based contrast agent enhances vascular bio-effects of 
1.09 MHz ultrasound on mouse intestine. Ultrasound Med Biol, 24, 1201-1208. 

121 



8 The safe use of contrast-enhanced diagnostic ultrasound 



Miller DL, Quddus J. 2000a. Sonoporation of monolayer cells by diagnostic ultrasound 

activation of contrast-agent gas bodies. Ultrasound Med Biol, 26, 661-667. 

Miller DL, Quddus J. 2000b. Diagnostic ultrasound activation of contrast agent gas bodies 

induces capillary rupture in mice. Proc Natl Acad Sci USA, 97, 10179-10184. 

Miller DL, Quddus J. 2001. Lysis and sonoporation of epidermoid and phagocytic 

monolayer cells by diagnostic ultrasound activation of contrast agent gas bodies. 

Ultrasound Med Biol, 27, 1107-1113. 

Miller DL, Li P, Gordon, D, Armstrong WF. 2005a. Histological characterization of 

microlesions induced by myocardial contrast echocardiography. Echocardiography, 22, 25-34. 

Miller DL, Li P, Dou C, Gordon D, Edwards CA, Armstrong WF. 2005b. Influence of 

contrast agent dose and ultrasound exposure on cardiomyocyte injury induced by 

myocardial contrast echocardiography in rats. Radiology, 237, 137-143. 

Miller DL, Driscoll EM, Dou C, Armstrong WF, Lucchesi BR. 2006. Microvascular 

permeabilization and cardiomyocyte injury provoked by myocardial contrast 

echocardiography in a canine model. } Am Coll Cardiol, 47, 1464-1468. 

Miller DL, Dou C, Wiggins RC, Wharram BL, Goyal M, Williams AR. 2007a. An in vivo 

rat model simulating imaging of human kidney by diagnostic ultrasound with gas-body 

contrast agent. Ultrasound Med Biol, 33, 129-135. 

Miller DL, Dou C, Wiggins RC. 2007b. Doppler mode pulse sequences mitigate glomerular 
capillary hemorrhage in contrast-aided diagnostic ultrasound of rat kidney. IEEE Trans 
Ultrason Ferroelectr Freq Control, 54, 1802-1810. 

Miller DL, Averkiou MA, Brayman AA, Everbach EC, Holland CK, Wible Jr JH, et al. 
2008a. Bioeffects considerations for diagnostic ultrasound contrast agents. / Ultrasound 
Med, 27, 611-632. 

Miller DL, Dou C, Wiggins RC. 2008b. Frequency dependence of kidney injury induced 
by contrast-aided diagnostic ultrasound in rats. Ultrasound Med Biol, 34, 1678-1687. 
Miller DL, Dou C, Wiggins RC. 2009. Glomerular capillary hemorrhage induced in rats by 
diagnostic ultrasound with gas-body contrast agent produces intratubular obstruction. 

Ultrasound Med Biol, 35, 869-877. 

Miller DL, Dou C, Wiggins RC. 2010a. Contrast-enhanced diagnostic ultrasound causes 
renal tissue damage in a porcine model. / Ultrasound Med, 29, 1391-1401. 
Miller DL, Dou C, Wiggins RC. 2010b. In vivo gas body efficacy for glomerular capillary 
hemorrhage induced by diagnostic ultrasound in rats. IEEE Trans BioMed Eng, 57, 167-174. 
Miller DL, Dou C, Lucchesi BR. 2011. Are ECG premature complexes induced by 
ultrasonic cavitation electrophysiological responses to irreversible cardiomyocyte injury? 
Ultrasound Med Biol, 37, 312-320. 

Molina CA, Ribo M, Rubiera M, Montaner J, Santamarina E, Delgado-Mederos R, 
et al. 2006. Microbubble administration accelerates clot lysis during continuous 2-MHz 
ultrasound monitoring in stroke patients treated with intravenous tissue plasminogen 
activator. Stroke, 37, 425-429. 

Mulvagh SL, Rakowski H, Vannan MA, Abdelmoneim SS, Becher H, Bierig SM, 
et al. 2008. American Society of Echocardiography Consensus Statement on the clinical 
applications of ultrasonic contrast agents in echocardiography. / Am Soc Echocardiogr, 
21, 1179-1201. 

Mychaskiw 2nd G, Badr AE, Tibbs R, Clower BR, Zhang JH. 2000. Optison (FS069) 
disrupts the blood-brain barrier in rats. Anesth Analg, 91, 798-803. 

122 



The safe use of contrast-enhanced diagnostic ultrasound 8 



Nanda NC, Schlief R, Goldberg BB (editors). 1997. Advances in Echo Imaging Using 
Contrast Enhancement. 2nd Edition. London, UK: Kluwer Academic Publishers, pp. 1-698. 
NCRP. 2002. Exposure Criteria for Medical Diagnostic Ultrasound: II. Criteria Based on 
All Known Mechanisms. Report no. 140. Bethesda, MD: National Council on Radiation 
Protection and Measurements, pp. 213-239. 

Okazaki J, Ishikura F, Asanuma T, Otani K, Beppu S. 2004. Premature ventricular 
contraction during myocardial contrast echocardiography: relationship with imaging 
method, acoustic power and dose of contrast agent. / Cardiol, 43, 69-74. 
Ostensen J, Hede R, Myreng Y, Ege T, Holtz E. 1992. Intravenous injection of Albunex 
microspheres causes thromboxane mediated pulmonary hypertension in pigs, but not in 
monkeys or rabbits. Acta Physiol Scand, 144, 307-315. 

Perren F, Loulidi J, Poglia D, Landis T, Sztajzel R. 2008. Microbubble potentiated 
transcranial duplex ultrasound enhances IV thrombolysis in acute stroke. / Thromb 
Thrombolysis, 25, 219-223. 

Piscaglia F, Bolondi L, Italian Society for Ultrasound in Medicine and Biology (SIUMB) 
Study Group on Ultrasound Contrast Agents. 2006. The safety of Sonovue in abdominal 
applications: retrospective analysis of 23188 investigations. Ultrasound Med Biol, 32, 
1369-1375. 

Porter TM, Smith DA, Holland CK. 2006. Acoustic techniques for assessing the Optison 
destruction threshold. / Ultrasound Med, 25, 1519-1529. 

Porter TR, Xie F. 2010. Myocardial perfusion imaging with contrast ultrasound. JACC 
Cardiovasc Imaging, 3, 176-187. 

Rasmussen H, Dirven HA, Grant D, Johnsen H, Midtvedt T. 2003. Etiology of cecal 
and hepatic lesions in mice after administration of gas-carrier contrast agents used in 
ultrasound imaging. Toxicol Appl Pharmacol, 188, 176-184. 

Rubiera M, Alexandrov AV. 2010. Sonothrombolysis in the management of acute ischemic 
stroke. Am } Cardiovasc Drugs, 10, 5-10. 

Schlachetzki F, Holscher T, Koch HJ, Draganski B, May A, Schuierer G, et al. 2002. 
Observation on the integrity of the blood-brain barrier after microbubble destruction by 
diagnostic transcranial color-coded sonography. / Ultrasound Med, 21, 419-429. 
Schlief R, Schurmann R, Balzer T, Zomack M, Niendorf H. 1993. Saccharide based contrast 
agents. In Advances in Echo Imaging Using Contrast Enhancement, Nanda NC, Schlief R 
(editors). Dordrecht, the Netherlands: Kluwer Academic Publishers, pp. 71-96. 
Schneider M, Arditi M, Barrau M, Brochot J, Broillet A, Bentrone R, et al. 1995. BR1: a new 
ultrasonographic contrast agent based on sulfur hexafluoride-filled microbubbles. Invest 
Radiol, 30, 451-457. 

Shah RR. 2005. Drugs, QTc interval prolongation and final ICH E14 guideline: an important 
milestone with challenges ahead. Drug Saf, 28, 1009-1028. 

Shi WT, Forsberg F, Tornes A, Ostensen J, Goldberg BB. 2000. Destruction of contrast 
microbubbles and the association with inertial cavitation. Ultrasound Med Biol, 26, 1009-1019. 
Shigeta K, Itoh K, Ookawara S, Taniguchi N, Omoto K. 2004. Endothelial cell injury and 
platelet aggregation induced by contrast ultrasonography in the rat hepatic sinusoid. 
/ Ultrasound Med, 23, 29-36. 

Shigeta K, Itoh K, Ookawara S, Taniguchi N, Omoto K. 2005. The effects of Levovist and 
DD-723 in activating platelets and damaging hepatic cells of rats. / Ultrasound Med, 24, 
967-974. 

123 



8 The safe use of contrast-enhanced diagnostic ultrasound 

Skyba DM, Price RJ, Linka AZ, Skalak TC, Kaul S. 1998. Direct in vivo visualization of 
intravascular destruction of microbubbles by ultrasound and its local effects on tissue. 
Circulation, 98, 290-293. 

Stride E, Saffari N. 2003. On the destruction of microbubble ultrasound contrast agents. 
Ultrasound Med Biol, 29, 563-573. 

Szebeni J, Alving CR, Rosivall L, Bunger R, Baranyi L, Bedocs P, et al. 2007. Animal models 
of complement-mediated hypersensitivity reactions to liposomes and other lipid-based 
nanoparticles. / Liposome Res, 17, 107-117. 

Thomsen HS, Muller RN, Mattrey RF (editors). 1999. Trends in Contrast Media. Berlin, 
Germany: Springer- Verlag, pp. 1-480. 

Unger E, Fritz T, McCreery T, Sahn D, Barrette T, Yellowhair D, et al. 1997. Liposomes 
as myocardial perfusion ultrasound contrast agents. In Ultrasound Contrast Agents, 
Goldberg BB (editor). London, UK: Martin Dunitz, pp. 57-74. 

Vancraeynest D, Havaux X, Pouleur AC, Pasquet A, Gerber B, Beauloye C, et al. 2006. 
Myocardial delivery of colloid nanoparticles using ultrasound-targeted microbubble 
destruction. Eur Heart ], 27, 237-245. 

Vancraeynest D, Kefer J, Hanet C, Fillee C, Beauloye C, Pasquet A, et al. 2007. Release of 
cardiac bio-markers during high mechanical index contrast-enhanced echocardiography 
in humans. Eur Heart ], 28, 1236-1241. 

Vancraeynest D, Havaux X, Pasquet A, Gerber B, Beauloye C, Rafter P, et al. 2009. 

Myocardial injury induced by ultrasound-targeted microbubble destruction: evidence for 

the contribution of myocardial ischemia. Ultrasound Med Biol, 35, 672-679. 

Van der Wouw P, Brauns AC, Bailey SE, Powers JE, Wilde AA. 2000. Premature ventricular 

contractions during triggered imaging with ultrasound contrast. / Am Soc Echocard, 13, 

288-294. 

Whellan DJ, Green CL, Piccini JP, Krucoff MW. 2009. QT as a safety biomarker in drug 
development. Clin Pharmacol Ther, 86, 101-104. 

Wible Jr JH, Galen KP, Wojdyla JK, Hughes MS, Klibanov AL, Brandenburger GH. 
2002. Microbubbles induce renal hemorrhage when exposed to diagnostic ultrasound in 
anesthetized rats. Ultrasound Med Biol, 28, 1535-1546. 

Williams AR, Wiggins RC, Wharram BL, Goyal M, Dou C, Johnson KJ, et al. 2007. Nephron 
injury induced by diagnostic ultrasound imaging at high mechanical index with gas body 
contrast agent. Ultrasound Med Biol, 33, 1336-1344. 

Wilson SR, Burns PN. 2010. Microbubble-enhanced US in body imaging: what role? 
Radiology, 257, 24-39. 



124 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 9 

Epidemiological prenatal 
ultrasound studies 



Kjell A. Salvesen 

Department of Obstetrics and Gynaecology, Clinical Sciences, 

Lund University, Lund, Sweden 



Summary 

• Systematic reviews of epidemiological studies have shown no association between 
prenatal ultrasound and adverse outcomes. 

• There is a weak statistical significant association between prenatal ultrasound and 
being non-right handed. 

• Most epidemiologic evidence derives from B-mode scanners in commercial use 
20-25 years ago, and acoustic outputs from modern scanners are higher. One must 
acknowledge that the available epidemiological data are limited. 



9.1 Introduction 

Ultrasound has an extraordinary safety record. It has been used in obstetrics for almost 
four decades with no proven harmful effects. However, absence of evidence of harm 
is not evidence of absence of harm. Thus, it is necessary to study the effect of prenatal 
ultrasound exposure on human populations directly before any definitive statements 
regarding risk can be made. 

Epidemiological studies may be divided into observational and experimental studies The best 
(see Figure 9.1). The observational studies are in turn subdivided into descriptive and 
analytical studies. Descriptive studies are suitable for generating new hypotheses about controlled trials 
associations between exposure and disease, whereas analytical studies are designed 
to test such hypotheses. The simplest descriptive study is the cross-sectional study, in 
which patients are examined only once. In a longitudinal study, the patients are followed 
over time, but the study is still classified as hypothesis generating. Analytical studies, on 
the other hand, have a prior hypothesis of a possible association between exposure and 
disease. They, in turn, are subdivided into cohort and case-control studies depending 
on whether the scientist starts out with a hypothesis about the exposure or about the 
disease. In the simplest case-control design, patients are examined only once, whereas 

125 



studies involve 
randomized 



9 Epidemiological prenatal ultrasound studies 



Observational 
studies 




Descriptive 
studies 




Analytical 
studies 



Cross-sectional 
studies 



Longitudinal 
studies 




Case-Control 
studies 



Cohort 
studies 



Experimental 
studies 



Natural 
experiments 



Randomized 
controlled trials 



Figure 9.1. Classification of epidemiological studies. 



in a cohort study it is necessary to follow patients over time. The randomized controlled 
trial is regarded as being the best way to examine possible cause-effect relationships in 
human populations. Evidence based medicine is usually derived from systematic review 
of data obtained from randomized controlled trials and/or observational studies. For 
interpretation of data derived from epidemiological studies, there is a hierarchy of studies 
based on study design and the quality of the research methods. Greatest value should be 
given to systematic reviews of randomized controlled trials, and less to cohort studies, 
case-control studies and other observational studies (in that order). 

Two systematic reviews of epidemiological studies of the safety of ultrasound in 
pregnancy have been published (Torloni et ah, 2009; Whitworth et ah, 2010). The 
literature was searched extensively by the authors of the ISUOG-WHO review (up to 
October 2007) (Torloni et ah, 2009) and the Cochrane review (up to September 2009) 
(Whitworth et ah, 2010). The Cochrane review (Whitworth et ah, 2010) included all 
registered published and ongoing randomized controlled trials and quasi-randomized 
trials, but no analytical studies. The ISUOG-WHO review (Torloni et ah, 2009) included 
16 randomized controlled trials, 13 cohort and 12 case-control studies published 

126 



Epidemiological prenatal ultrasound studies 9 



between January 1950 and October 2007 that assessed any type of short- and long-term 
effects of at least one exposure to ultrasound during pregnancy. The authors screened 
6716 titles/abstracts and included 61 papers in the systematic review. The outcomes 
assessed included maternal outcomes, perinatal outcomes, childhood growth, 
neurological development and school performance, non-right handedness, childhood 
malignancies, intellectual performance and mental diseases after childhood (Torloni 
et al, 2009). 



This chapter adopts the main results from the ISUOG-WHO systematic review and 
is expanded with the results from two epidemiological studies published after 2007 
(Stalberg et al, 2009; Heikkila et al, 2011). There is, however, one caveat about the data 
from epidemiological studies. The acoustic outputs from modern devices have increased 
10-15-fold during the last decades (Henderson et al., 1997), and most epidemiologic 
evidence derives from B-mode scanners in commercial use 20-25 years ago. If adverse 
effects of ultrasound during pregnancy are dose dependent, one must acknowledge that 
the available epidemiological data are limited. 



Evidence based 
medicine is 
based on 
systematic 
reviews 



9.2 Adverse perinatal outcomes 



9.2.1 Birth weight 

The question of whether ultrasound exposure in utero leads to reduced birth weight has 
probably been given more attention than any other perinatal outcome. This may be due 
to the existence of such an effect in some animal models, and/or because it is relatively 
quick and easy to measure. In 9 controlled trials involving over 35,000 women (Duff 1993; 
Ewigman et al, 1993; Doppler French study group, 1997; Geerts et al, 1996; Kieler et al, 
1998c; Mason et al, 1993; Newnham et al, 2004; Omtzigt et al, 1994; Secher et al, 1986) and 
4 cohort studies with another 2000 women (Bellieni et al, 2005; Geerts et al, 2004; Smith, 
1984; Stark et al, 1984) exposure to ultrasound during pregnancy did not significantly 
influence the mean birth weight of the offspring. Similarly pooled odds ratios (ORs) from 
controlled trials and cohort studies do not seem to increase the proportions of low-birth 
weight children (Torloni et al, 2009). 



There is no 
convincing 
evidence that 
ultrasound 
exposure in 
utero affects 
birth weight 



9.2.2 Perinatal mortality 

Perinatal mortality has been regarded as the most important outcome to study in 
controlled trials related to introducing routine ultrasound screening into antenatal care. 
Sceptics about new technology argue that if one cannot demonstrate a reduction in 
perinatal mortality, pregnant women should not be offered routine ultrasound screening. 
From a safety perspective it is just as important to demonstrate that ultrasound exposure 
in pregnancy does not increase perinatal mortality. 

Perinatal mortality has been studied in 13 controlled trials (Bakketeig et al, 1984; Crowther 
et al, 1999; Davies et al, 1992; Duff, 1993; Eik-Nes et al, 2000; Ewigman et al, 1990; 
Ewigman et al, 1993; Doppler French study group, 1997; Geerts et al, 1996; Omtzigt et al, 
1994; Saari-Kemppainen et al, 1990; Secher et al, 1986; Waldenstrom, 1988). Among 46,553 

127 



9 Epidemiological prenatal ultrasound studies 



women included in the controlled trials there was a non-significant 14% reduction in 
perinatal mortality in the ultrasound group [OR 0.86, 95% confidence interval (CI) 
0.70-1.07] (Torloni et al., 2009). A similar non-significant reduction in perinatal mortality 
(OR 0.89, 95% CI 0.75-1.07) was found in a cohort study of approximately 210,000 Swedish 
women (Sylvan et al., 2005). 



9.2.3 Other perinatal outcomes 



There is no 
evidence that 
ultrasound 
exposure 
in utero 
affects 
perinatal 
morbidity or 
mortality 



Obstetricians and neonatologists commonly assess neonatal morbidity and labour 
outcomes from the rates of preterm birth, low Apgar score at 5min and admission to 
neonatal intensive care unit (NICU). Preterm birth has been studied in 8 controlled trials 
involving 34,000 women, and no association was found (Torloni et al., 2009). Similarly, 
for low Apgar score at 5 min (12 controlled trials involving 22,000 women) and admission 
to NICU (13 controlled trials involving 33,000 women) there were no adverse effects of 
ultrasound during pregnancy (Torloni et al., 2009). 



9.3 Childhood malignancies 



There is no 
evidence that 
ultrasound 
exposure 
in utero is 
associated 
with childhood 
malignancies 



When the outcome under study is rare, such as with childhood malignancies, any 
approach other than the case-control design is unsuitable. For childhood malignancies 
there are data from 8 studies including more than 14,000 children (Bunin et al., 1994; 
Cartwright et al., 1984; Wilson and Waterhouse, 1984; Naumburg et al., 2000; Shu 
et al., 1994, 2002; Sorahan et al, 1995; Stalberg et al, 2008). No associations between 
prenatal ultrasound exposure and childhood malignancies have been found (Torloni 
et al, 2009). Analyses have also been carried out for subgroups of malignancies. For 
leukaemia there are data from 5 case-control studies with 6334 children (Cartwright 
et al, 1984; Wilson and Waterhouse, 1984; Naumburg et al, 2000; Shu et al, 1994, 2002), 
and for central nervous system (CNS) tumours there are data from 3 case-control 
studies with 1909 children (Bunin et al, 1994; Cartwright et al, 1984; Stalberg et al, 
2008). No associations have been found between prenatal ultrasound and leukaemia 
or CNS tumours. 



9.4 Neurological development, dyslexia and speech 
development 

These outcomes have been studied in two controlled trials with more than 5200 children 
(Kieler et al, 1998a; Salvesen et al, 1992; Salvesen et al, 1994), one case-control study with 
214 children (Campbell et al, 1993) and one cohort study with 806 children (Stark et al, 
1984). The two observational studies (Campbell et al, 1993; Stark et al, 1984) published 
possible associations between prenatal ultrasound and dyslexia (Stark et al, 1984) and 
delayed speech development (Campbell et al, 1993). However, the controlled trials 
found no statistically significant associations between prenatal ultrasound and a long 
list of outcomes, such as dyslexia, delayed speech, stuttering, poor vocabulary, referral to 
speech therapist, various neurological tests, impaired vision or impaired hearing (Torloni 
et al, 2009). Overall the results suggest that it is unlikely that prenatal ultrasound can 
"cause harm to the developing foetal brain". 

128 



Epidemiological prenatal ultrasound studies 9 



9.5 School performance 

School performance has been studied in two controlled trials with almost 6500 children 
(Salvesen et ah, 1992; Stalberg et ah, 2009). Salvesen et al. followed up children at 8-9 years 
and found no association between prenatal ultrasound and poor school performance, 
including arithmetic scores, spelling, reading comprehension and oral reading (Salvesen 
et ah, 1992). Stalberg et al. followed up children at 15-16 years and found no statistically 
significant differences in school performance for boys or girls according to randomization 
or exposure to ultrasound in the second trimester (Stalberg et al., 2009). Compared to 
those who were unexposed, boys exposed to ultrasound had a tendency towards lower 
mean school grades in general and in physical education, but the differences did not 
reach statistical significance (Stalberg et ah, 2009). 



9.6 Intellectual performance and mental disease in 
adult life 

In a Swedish cohort study of 7999 prenatally ultrasound exposed and 197,829 unexposed 
men aged 18 years, there was an increased risk of subnormal intellectual performance 
(OR 1.19, 95% CI 1.12-1.27) among exposed men (Kieler et ah, 2005). However, this 
association was probably confounded by sociogeographical factors, and within pairs 
of brothers there was no association between ultrasound exposure and intellectual 
performance. Thus, the authors concluded that "the study failed to demonstrate a clear 
association between ultrasound and intellectual performance" (Kieler et ah, 2005). In 
another Swedish cohort study of 370,945 individuals there was no association between 
prenatal ultrasound and schizophrenia (OR 1.47, 95% CI 0.99-2.16) or other psychoses 
(OR 1.03, 95% CI 0.80-1.33) (Stalberg et ah, 2007). 

9.7 Handedness 

There is, however, one statistically significant association between prenatal ultrasound 
exposure and behaviour that holds up through all epidemiological studies and systematic 
reviews. The controversy is handedness. 



There is no 
evidence that 
ultrasound 
exposure in 
utero affects 
neurological 
development, 
dyslexia, speech 
development, 
school 

performance, 
intellectual 
performance 
or adult mental 
disease 



The first meta-analysis demonstrating an association between ultrasound and non-right 
handedness was published in 1999 (Salvesen and Eik-Nes, 1999). There was no statistically 
significant difference in the prevalence of non-right handedness between ultrasound 
screened children and controls (OR 1.13, 95% CI 0.97-1.32), but there was a difference in a 
subgroup analysis of boys (OR 1.26, 95% CI 1.03-1.34) (Salvesen and Eik-Nes, 1999). In the 
most recent Cochrane review (Whitworth et ah, 2010) a conservative approach towards 
subgroup analyses is advocated. The Cochrane review reports no association between 
ultrasound and non-right handedness in an intention-to-treat analysis of all children 
(OR 1.12, 95% CI 0.92-1.36) and abstains from doing a gender specific subgroup analysis. 
The ISUOG-WHO review (Torloni et ah, 2009) has adopted a less conservative analytical 
approach and included two randomized trials (Salvesen et ah, 1993a,b; Kieler et ah, 1998b) 
and two cohort studies (Kieler et ah, 2001, 2002). The ISUOG-WHO review confirms the 
findings from the first meta-analysis (Salvesen and Eik-Nes, 1999), and adds: "When boys 

129 



9 Epidemiological prenatal ultrasound studies 





Study or Subgroup 




und | 


Control 


Weight 


Peto Odds Ratio 




Events 


Total 


Events 


Total 


Peto, Fixed, 95KCI 


0 Finland 


353 


2112 


300 


2038 


45.9% 


1.16 (0.98. 1.37] 


0 Norway 


162 


sei 


120 


802 


19.5% 


1.31 [1.02, 1.70J 


0 Sweden 


253 


1544 


240 


1508 


34.5% 


1.04 [0.8S, 1.26] 




Total (95% CI) 




4517 




4348 


100.0% 


1.14 [1.02, 1.28] 




Total events 


768 




660 









Peto Odds Ratio 


Peto, Fixed, 95KCI 




















0.5 0.7 

Favours experimental 


1.5 2 

Favours control 



Figure 9.2. Non-right handedness among all children compared according to the randomized 
groups from three follow-up studies of randomized controlled trials (Heikkila et al., 2011; 
Salvesen eta/., 1993a; Kieler et a/., 1998c). 



Table 9.1. Non-right handedness (NRH) according to randomized groups 





Ultrasound group (n) 


Controls (n) 






Study 


NRH 


Total 


NRH 


Total 


Weight 
(%) 


OR 

(95% CI) 


Heikkila 
et al, 2011 


353 


2112 


300 


2038 


45.9 


1.16 (0.98-1,37) 


Salvesen 
et al, 1993a 

Kieler et al, 
1998b 


162 
253 


861 
1544 


120 
240 


802 
1508 


19.5 
34.5 


1.31 (1.02-1.70) 
1.04 (0.85-1.26) 


Total 


768 


4517 


660 


4348 


100 


1.14 (1.02-1.28) 



were considered separately, there is a weak association between ultrasound screening 
and being non-right handed, both in the randomized trials (OR 1.26, 95% CI 1.03-1.34) 
and in the cohort studies (OR 1.17, 95% CI 1.07-1.28)" (Torloni et al, 2009). 

A follow-up study of a Finnish randomized trial was published in 2011 (Heikkila 
et al, 2011). At first glance this study appeared reassuring since there is no difference in 
non-right handedness between ultrasound screened children and controls (OR 1.16, 95% 
CI 0.98-1.37), nor in a subgroup analysis of boys (OR 1.12, 95% CI 0.89-1.41) (Heikkila 
et al., 2011). However, when the results of the Finnish trial were included in a new 
meta-analysis of three randomized controlled trials, there was a statistically significant 
increased prevalence of non-right handedness in ultrasound screened children compared 
with controls (OR 1.14, 95% CI 1.02-1.28) (Salvesen, 2011) (Figure 9.2 and Table 9.1). The 
results in subgroups according to gender were consistent with overall results with no 
significant differences between boys and girls, but among boys the association became 
stronger when an exploratory analysis according to ultrasound exposure before 19-22 
weeks was done (OR 1.30, 95% CI 1.10-1.53) (Salvesen, 2011). 



Prenatal 
ultrasound 
is associated 
with 
non-right 
handedness 
in children 



A final conclusion of a possible association between prenatal ultrasound and handedness 
cannot yet be drawn. According to the data presented in the Cochrane review (With worth, 
2010), there is no statistically significant association between prenatal ultrasound and 
left-handedness. However, in the most recent meta-analysis including three randomized 
controlled trials there was a statistically signifi ant association (Salvesen, 2011). If 
results from cohort studies are accepted (Kieler et al., 2001, 2002), as was done in the 

130 



Epidemiological prenatal ultrasound studies 9 



ISUOG-WHO review, the strength of the association was similar for randomized 
trials and cohort studies (Torloni et al., 2009). Thus, the conclusion must be that five 
epidemiological studies report a 15% increase in the likelihood of sinistrality (in particular 
among males), and no other epidemiological evidence contradicts this association. 

The discussion of prenatal ultrasound and left-handedness is complex and will not be 
extended here. An editorial explores this issue in detail (Salvesen, 2002). A statistical 
association between ultrasound and left-handedness should not lead to the conclusion that 
ultrasound causes harm to the developing brain. The current biological understanding 
of handedness is limited and partly contradictory to the epidemiological evidence 
(Salvesen, 2002). 

9.8 Conclusion 

Epidemiological studies have demonstrated no confirmed associations between prenatal 
ultrasound and adverse perinatal outcomes, childhood malignancies, neurological 
development, dyslexia, speech development, school performance, intellectual 
performance and adult mental disease. However, there is a weak statistical significant 
association between prenatal ultrasound and being non- right handed. This does not mean 
that there must be a causal relationship. We will have to live with uncertainty regarding 
ultrasound safety in the years to come. 

References 

Bakketeig L, Jacobsen G, Brodtkorb CJ, Eriksen BC, Eik-Nes SH, Ulstein MK, et al. 1984. 
Randomised controlled trial of ultrasonographic screening in pregnancy. Lancet, 2, 207-21 1 . 
Bellieni CV, Buonocore G, Bagnoli F, Cordelli DM, Gasparre O, Calonaci F, et al. 2005. Is an 
excessive number of prenatal echographies a risk for fetal growth? Early Hum Dev, 81, 689-693. 
Bunin GR, Buckley JD, Boesel CP, Rorke LB, Meadows AT. 1994. Risk factors for astrocytic 
glioma and primitive neuroectodermal tumor of the brain in young children: a report 
from the Children's Cancer Group. Cancer Epidemiol Biomark Prev, 3, 197-204. 
Cartwright RA, McKinney PA, Hopton PA, Birch JM, Hartley AL, Mann JR, et al. 1984. 
Ultrasound examination in pregnancy and childhood cancer. Lancet, 2, 999-1000. 
Campbell JD, Elford RW, Brant RF. 1993. Case-Control study of prenatal ultrasonography 
exposure in children with delayed speech. Can Med Assoc }, 149, 1435-1440. 
Crowther CA, Kornman L, O'Callaghan S, George K. Furness M, Wilson K. 1999. Is an 
ultrasound assessment of gestational age at the first antenatal visit of value? A randomised 
clinical trial. Br J Obstet Gynaecol, 106, 1273-1279. 

Davies J, Gallivan S, Spencer J. 1992. Randomised controlled trial of Doppler ultrasound 
screening of placental perfusion during pregnancy. Lancet, 340, 1299-1303. 
Doppler French Study Group. 1997. A randomised controlled trial of Doppler ultrasound 
velocimetry of the umbilical artery in low risk pregnancies. Br ] Obstet Gynaecol, 104, 419^24. 
Duff GB. 1993. A randomized controlled trial in a hospital population of ultrasound 
measurement screening for the small for dates baby. Aust NZJ Obstet Gynaecol, 33, 374-378. 
Eik-Nes SH, Salvesen KA, Okland O, Vatten LJ. 2000. Routine ultrasound fetal examination in 
pregnancy: the "Alesund" randomized controlled trial. Ultrasound Obstet Gynecol, 15, 473^78. 

131 



9 Epidemiological prenatal ultrasound studies 



Ewigman BG, Crane JP, Frigoletto FD, LeFevre M, Bain RP, McNellis D. 1993. Effect 
of prenatal ultrasound screening on perinatal outcome. RADIUS Study Group. N Engl 
J Med, 329, 821-827. 

Ewigman BG, LeFevre M, Hesser J. 1990. A randomized trial of routine prenatal 
ultrasound. Obstet Gynecol, 76, 189-194. 

Geerts L, Theron AM, Grove D, Theron GB, Odendaal HJ. 2004. A community-based 
obstetric ultrasound service, Int J Gynaecol Obstet, 84, 23-31. 

Geerts L, Brand E, Theron G. 1996. Routine ultrasound examinations in South Africa: cost 
and effect on perinatal outcome - a prospective randomised controlled trial. Br ] Obstet 
Gynaecol, 103, 501-507. 

Henderson J, Whittingham TA, Dunn T. 1997. A review of the acoustic output of modern 
diagnostic ultrasound equipment. BMUS Bull, 10-14. 

Heikkila K, Vuoksimaa E, Oksava K, Saari-Kemppainen A, Iivanainen M. 2011. 
Handedness in the Helsinki ultrasound trial. Ultrasound Obstet Gynecol, 37, 638-642. 
doi: 10.1002/uog.8962. 

Kieler H, Hellberg D, Nilsson S, Waldenstrom U, Axelsson 0. 1998c. Pregnancy outcome among 
non-participants in a trial on ultrasound screening. Ultrasound Obstet Gynecol, 11, 104-109. 
Kieler H, Haglund B, Cnattingius S, Palmgren J, Axelsson O. 2005. Does prenatal 
sonography affect intellectual performance? Epidemiology, 16, 304-310. 
Kieler H, Ahlsten G, Haglund B, Salvesen K, Axelsson 0. 1998a. Routine ultrasound screening 
in pregnancy and the children's subsequent neurologic development. Obstet Gynecol, 91, 750-756. 
Kieler H, Axelsson O, Haglund B, Nilsson S, Salvesen K. 1998b. Routine ultrasound 
screening in pregnancy and the children's subsequent handedness. Early Hum Dev, 50, 
233-245. 

Kieler H, Cnattingius S, Haglund B, Palmgren J, Axelsson O. 2001. Sinistrality - a side- 
effect of prenatal sonography: a comparative study of young men. Epidemiology, 12, 618-623. 
Kieler H, Cnattingius S, Palmgren J, Haglund B, Axelsson O. 2002. First trimester 
ultrasound scans and left-handedness. Epidemiology, 13, 370. 

Mason GC, Lilford RJ, Porter J, Nelson E, Tyrell S. 1993. Randomised comparison 
of routine versus highly selective use of Doppler ultrasound in low risk pregnancies. 
Br J Obstet Gynaecol, 100, 130-133. 

Naumburg E, Bellocco R, Cnattingius S, Hall P, Ekbom A. 2000. Prenatal ultrasound 
examinations and risk of childhood leukaemia: case-control study. Br Med ], 320, 282-283. 
Newnham J, Doherty DA, Kendall GE, Zubrick SR, Landau LL, Stanley FJ. 2004. Effects 
of repeated prenatal ultrasound examinations on childhood outcome up to 8 years of age: 
follow-up of a randomised controlled trial. Lancet, 364, 2038-2044. 

Omtzigt AM, Reuwer PJ, Bruinse HW. 1994. A randomized controlled trial on the clinical 
value of umbilical Doppler velocimetry in antenatal care. Am J Obstet Gynecol, 170, 
625-634. 

Saari-Kemppainen A, Karjalainen O, Ylostalo P, Heinonen OP. 1990. Ultrasound screening 
and perinatal mortality: controlled trial of systematic one-stage screening in pregnancy. 
Lancet, 336, 387-391. 

Salvesen KA, Vatten LJ, Bakketeig LS, Eik-Nes SH. 1994. Routine ultrasonography in utero 
and speech development. Ultrasound Obstet Gynecol, 4, 101-103. 

Salvesen KA, Bakketeig LS, Eik-Nes SH, Undheim JO, Okland O. 1992. Routine 
ultrasonography in utero and school performance at age 8-9 years. Lancet, 339, 85-89. 

132 



Epidemiological prenatal ultrasound studies 9 



Salvesen KA, Vatten LJ, Eik-Nes SH, Hugdahl K, Bakketeig LS. 1993a. Routine 
ultrasonography in utero and subsequent handedness and neurological development. 
Br Med J, 307,159-164. 

Salvesen KA, Eik-Nes SH, Vatten LJ, Hugdahl K, Bakketeig LS. 1993b. Routine ultrasound 
scanning in pregnancy - Authors' reply. BM], 307, 1562. 

Salvesen KA, Eik-Nes SH. 1999. Ultrasound during pregnancy and subsequent childhood 
non-right handedness: a meta-analysis. Ultrasound Obstet Gynecol, 13, 241-246. 
Salvesen KA. 2002. Ultrasound and left-handedness: a sinister association? Ultrasound 
Obstet Gynecol, 19, 217-221. 

Salvesen KA. 2011. Ultrasound in pregnancy and non-right handedness - meta-analysis of 
randomized trials. Ultrasound Obstet Gynecol, 38, 267-271. 

Secher NJ, Hansen PK, Lenstrup C, Eriksen PS. 1986. Controlled trial of ultrasound 
screening for light for gestational age (LGA) infants in late pregnancy. Eur ] Obstet Gynecol 
Re-prod Biol, 23, 307-313. 

Shu XO, Jin F, Linet MS, Zheng W, Clemens J, Mills J, et al. 1994. Diagnostic X-ray and 
ultrasound exposure and risk of childhood cancer. Br J Cancer, 70, 531-536. 
Shu XO, Potter JD, Linet MS, Severson RK, Han D, Kersey JH, et al. 2002. Diagnostic 
X-rays and ultrasound exposure and risk of childhood acute lymphoblastic leukemia by 
immunophenotype. Cancer Epidemiol Biomarkers Prev, 11, 177-185. 

Smith CB. 1984. Birth weights of fetuses exposed to diagnostic ultrasound. / Ultrasound 
Med, 3, 395-396. 

Sorahan T, Lancashire R, Stewart A, Peck I. 1995. Pregnancy ultrasound and childhood 
cancer: a second report from the Oxford Survey of Childhood Cancers. Br } Obstet Gynaecol, 
102, 831-832. 

Stark C, Orleans M, Haverkamp A, Murphy J. 1984. Short and long term risks after 

exposure to diagnostic ultrasound in utero. Obstet Gynecol, 63, 194-200. 

Stalberg K, Haglund B, Axelsson O, Cnattingius S, Pfeifer S, Kieler H. 2008. Prenatal ultrasound 

and the risk of childhood brain tumour and its subtypes. Br } Cancer, 98, 1285-1287. 

Stalberg K, Haglund B, Axelsson O, Cnattingius S, Hultman CM, Kieler H. 2007. Prenatal 

ultrasound scanning and the risk of schizophrenia and other psychoses. Epidemiology, 18, 577-582. 

Stalberg K, Axelsson O, Haglund B, Hultman CM, Lambe M, Kieler H. 2009. Prenatal 

ultrasound exposure and children's school performance at age 15-16: follow-up of a 

randomized controlled trial. Ultrasound Obstet Gynecol, 34, 297-303. 

Sylvan K, Ryding EL, Rydhstrom H. 2005. Routine ultrasound screening in the third 

trimester: a population based study. Acta Obstet Gynecol Scand, 84, 1154-1158. 

Torloni MR, Vedmedovska N, Merialdi M, Betran AP, Allen T, Gonzalez R, et al. 2009. 

Safety of ultrasonography in pregnancy: WHO systematic review of the literature and 

meta-analysis. Ultrasound Obstet Gynecol, 33, 599-608. 

Waldenstrom U, Axelsson O, Nilsson S, Eklund G, Fall O, Lindeberg S, et al. 1988. Effects 
of routine one-stage ultrasound screening in pregnancy: a randomised controlled trial. 
Lancet, 2, 585-588. 

Whitworth M, Bricker L, Neilson JP, Dowswell T. 2010. Ultrasound for fetal assessment in 
early pregnancy. Cochrane Database Syst Rev, (4), Art. No.: CD007058. doi: 10.1002/14651858. 
CD007058.pub2. 

Wilson LMK, Waterhouse JA. 1984. Obstetric ultrasound and childhood malignancies. 
Lancet, 2, 997-999. 

133 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 10 

Safety standards and regulations: 
the manufacturers' responsibilities 

Francis A. Duck 

University of Bath, Bath, UK 



Summary 

• All medical diagnostic ultrasound equipment must be manufactured to conform to 
safety standards and regulations. 

• Separate regulations exist in Europe, in the USA and internationally. 

• The International Electrotechnical Commission sets international safety standards, 
including maximum allowable transducer surface temperatures. 

• The United States Food and Drug Administration (FDA) sets upper limits for 
ultrasound exposure. 

• In order to use the highest output levels allowed by the FDA, manufacturers must 
display safety indices: the mechanical index and the thermal index. 

• In Europe, the Medical Devices Directive requires manufacturers to demonstrate 
both safety and measurement accuracy. 



10.1 Introduction 



Ultrasound 
safety is 
subject to 
national and 
international 
standards and 
regulations 



The design and manufacture of all medical equipment, including diagnostic ultrasound 
equipment, is subject to a number of regulatory controls and standards that are intended 
to ensure that it may be operated safely. The ultrasound equipment industry is now 
truly international, and as a result the most important regulations are those that have 
international status, or have an equivalent impact. The three sources of regulations and 
standards which have effect in the UK are: 



1. Standards of the International Electrotechnical Committee (IEC). IEC standards are 
generally accepted in full as European Standards and also as British Standards. 

2. Standards and regulations used in the USA, including those of the US Government 
Department of Health and Human Services, Food and Drug Administration (FDA). 

3. European directives. 



134 



Safety standards and regulations: the manufacturers' responsibilities 10 



Table 10.1. The main purposes of some key regulations and standards for diagnostic ultra- 
sound safety. 



Body Regulation or standard Purpose 



FDA 


510(k) (FDA, 1997) 


To limit the maximum allowed 
acoustic output 


AIUM/NEMA 


"Output Display Standard" 
(AIUM/NEMA, 1998) 


To define methods for 
calculation and display of 
safety indices 


IEC 


IEC60601 Part 1 (IEC, 2005) 


To regulate for the safe design 
and manufacture of all medical 
equipment, including thermal, 
electrical and mechanical 
aspects 


IEC 


IEC60601 Part 2-37 (IEC, 2001) 


To define display of safety 
indices, and to limit transducer 
surface temperature 


IEC 


IEC62359 (IEC, 2006) 


To define safety indices 


EC 


MDD (European Communities, 
1993) 


To set general requirements 
for safety and measurement 
precision for all medical 
devices 



Each of the regulations has a particular primary purpose, which generally distinguishes it 
from the others. These purposes are set out in Table 10.1 for some of the more important 
of them. The regulations are referred to by their commonly-used abbreviations. Details 
are given later in the text. Inspection of Table 10.1 makes clear that there is only one 
regulation that serves to limit the ultrasound output that transducers may emit, and that 
is due to the FDA in the USA. Further regulations exist which serve as general standards 
for safety for all electro-medical equipment, of which ultrasound diagnostic equipment is 
one part, but which equally apply to such devices. 



Each safety 
regulation or 
standard has 
a different 
purpose 



10.2. International Standards from the International 
Electrotechnical Commission 



10.2.1 IEC60601 Part 1: General safety requirements 

The general requirements for the safe design of electro-medical equipment are set out 
in IEC60601 Part 1 (IEC, 2005). Contained within this standard are requirements for the 
safe design of mechanical parts, for electrical safety and for thermal safety. Of specific 
importance are the electrical earth leakage current associated with an ultrasound 
transducer (up to 1 kHz), which for normal condition use is 0.5 mA, and the patient 
leakage current, limited to 100 mA for general use and to 10 mA for intra-cardiac use. The 



All ultrasound 
scanners 
also must 
conform to 
electrical and 
thermal safety 
standards 



135 



10 Safety standards and regulations: the manufacturers' responsibilities 



safe management of heated parts that may make contact with the body is also considered. 
Particular specifications for ultrasound transducers are discussed below. 



IEC60601-2- 
37 sets limits 
to transducer 

surface 
temperatures, 
and specifies 
use of safety 

indices 



10.2.2 IEC60601 Part 2-37: Ultrasound diagnostic and monitoring 
equipment 

The IEC standard for diagnostic and monitoring equipment, IEC60601-2-37 (IEC, 2001), 
sets some further requirements over those set in IEC60601 Part 1. It specifies how a 
user shall be informed about potential hazard, through displayed indices related to 
exposure and safety. This approach to protection casts a responsibility on the user, 
requiring for its success an appropriate level of training and competence for all 
practitioners using ultrasound diagnostic equipment. The philosophy and definitions 
for index display are identical to those developed in the USA in the early 1990s, and 
first published in the so-called "Output Display Standard" (ODS; AIUM/NEMA, 1998) 
details of which are given in section 10.3. The methods for determination of the safety 
indices are also published in IEC62359 (IEC, 2006). IEC60601-2-37 also places limits 
on the allowed temperature of the surface of the transducer (Table 10.2). In air this 
is 50 °C, considerably higher than the temperature of 43 °C allowed in contact with 
tissue. Higher increases in temperature are allowed for transducers in contact with the 
skin than for those used internally, such as transvaginal or intra-rectal probes. These 
temperatures are not only of theoretical interest, since many scanners can now drive 
transducers at levels that approach them. 



FDA 
regulations 
are very 
influential 
in setting 
ultrasound 
output levels 



10.3. Regulations and standards from the USA 

10.3.1 FDA regulations (510k) 

The USA operates its own regulatory structure, through its FDA. These regulations have 
until now had world-wide impact, because they control the sale of ultrasound equipment 
in the USA, and the actions of US manufacturers elsewhere in the world. In spite of the 
emergence of stronger international and, now, European regulations, it seems likely that 
the FDA regulations will continue to influence manufacturers for a considerable time. 
The current situation in the USA is set out below. 



Table 10.2. Limits on surface temperature and surface temperature rise specified by IEC60601- 
2-37 Amendment 1 (IEC, 2001). 

I I On tissue On tissue 

(external use) (internal use) 

Maximum 5Q 43 ^ 

temperature (°C) 



Maximum 

temperature rise (°C) 



136 



Safety standards and regulations: the manufacturers' responsibilities 10 



Within the US Department of Health and Human Services, the FDA receives applications 
for "market approval" to sell equipment, through a process known as "510(k)" from the 
code used on earlier guidance documents issued by the FDA. The most recent guidance 
document was issued in 1997 (FDA, 1997). Two "tracks" are defined. These are known as 
Track 1 and Track 3. A very large proportion of modern equipment is designed according 
to requirements of Track 3, and so this is given emphasis here. Track 3 gives the provision 
for exploiting higher time-averaged intensities than those that were available under 
Track 1. These are allowed provided that the equipment has "real-time labelling" of 
exposure in accordance with the ODS published by AIUM/NEMA (American Institute of 
Ultrasound in Medicine/National Electrical Manufacturers' Association) (AIUM/NEMA, 
1998). In order to comply, values of mechanical index (MI) and thermal index (TI) must be 
displayed to the user. Details of the requirements of the ODS are given below. 



The FDA sets 
maximum 
allowed levels 
for output 
exposure. 
The highest 
intensities 
are allowed 
only if safety 
indices can be 
displayed 



The upper limits set by the FDA for Track 3 clearance are set out in Table 10.3. Two 
categories of use are defined: one small category is for equipment that is designed for use 
exclusively for ophthalmology, and the second is for equipment for all other applications, 
including obstetrics. Limiting values are given for four exposure quantities: derated (i.e. 
estimated in situ) spatialpeak, temporal-average intensity (I ^ d ), derated spatialpeak, 
pulse-average intensity (J sppad ), MI and TI. (The suffix d is used here to indicate de-rating.) 
Definitions of the two safety indices, MI and TI, are given below. 



FDA limits are 
the same for 
all clinical uses 
except for 
ophthalmology 



The limiting values for pulse-average intensity and MI are not independent: provided 
that either one of them lies below the specified limit then the other is allowed to exceed 
it. However, the limits TI = 1.9 and 7 sppad = 190 Wcirr 2 are deemed to be approximately 
equivalent. TI is used to control output only for ophthalmic applications. For any other 
application, manufacturers are required to provide the FDA with a justification for any 
use when the TI exceeds 6.0, but no absolute limiting value is specified. 



For all quantities the value is not that measured in water but the "estimated in situ" 
value. This is derived by de-rating the free-field value by 0.3 dB cm -1 MHz 1 . The position 
of which each quantity is to be calculated is where a derated value is maximum; for 
time-averaged intensity it is where I . is greatest; for MI it is where the derated pulse- 
intensity integral is greatest. 



FDA sets limits 
using estimated 
in situ values. 
These are 
lower than 
the equivalent 
values in water 



Table 10.3. The limits set by the US FDA on acoustic exposure allowed under "Track 3". At 
least one of Ml and / . must be lower than the specified limit, but not necessarily both 

sppa,d 1 J 

(FDA, 1997). 



Derated / sptM , 
(mWcrrr 2 ) 



Derated /. 
(Wcirr 2 ) 



sppa,d 



Ml 



TI 



All applications 
except ophthalmology 



720 



190 



1.9 



(6.0)* 



Ophthalmology 



51) 



Not specified 



0.23 



1.0 



*NB: The value of TI = 6 is not an upper limit for non-ophthalmic applications. Manufacturers are required to 
give a justification if conditions exist for which the TI exceeds 6.0. 



137 



10 Safety standards and regulations: the manufacturers' responsibilities 



The potential 
for higher 
output levels 
is particularly 
important 
in obstetric 
scanning 



The changes that were made to the FDA regulations in the early 1990s have meant that 
higher exposures, commonly associated with Doppler applications, may now exist on 
equipment intended for obstetric applications, for which the allowed upper limit for 
time-average intensity has increased about eight times. 

Values of other quantities also have to be submitted to the FDA. These include total acoustic 
power, centre frequency, beam dimensions at the focus, position of the measurement, 
pulse repetition frequency and focal length. Tables giving values for these quantities are 
often included with the technical documentation supplied with the scanner. 



10.3.2. AIUM/NEMAODS 

Definition of Ml As noted above, FDA Track 3 requires conformance with the AIUM/NEMA ODS (AIUM/ 
NEMA, 1998). The quantities in this standard are derived by calculation from measured 
acoustic pressure, power, intensity and frequency, and are intended to give guidance as 
to the possibility that the operating conditions might cause heating or mechanical effects. 
The following definitions also form part of IEC standards 60601-2-37 and 62359 (IEC, 
2001, 2006). 



The MI is calculated from the peak derated rarefaction pressure, p r . in MPa: 

p 

Ml=-£i 

41 

where/is the centre frequency in MHz; MI is given without units. 



(10.1) 



Definition 
of Tl 



TI is defined as W/W deg , where W is the acoustic power and W ieg is the power required to 
cause a maximum temperature rise of 1 °C under identical conditions of transducer, beam 
and tissue. TI is therefore taken to indicate broadly a reasonable estimate of the greatest 
steady-state temperature rise in degrees Celsius. 



There are three 
TI categories, for 
soft tissue, for 
bone-at-focus 
and for cranial 
bone 



There are a number of different formulae for TI. These depend on the expected target, 
on transducer size, and on whether the beam is scanned or unscanned (see Figure 10.1). 
There are three TI categories related to three targets: soft tissue (TIS), bone-at-focus (TIB) 
and adult cranial (TIC). 



Formulae for 
soft-tissue TI 
depend on 
frequency: those 
for bone do not. 
Both depend on 
acoustic power 



The formulae for TI can be simplified and summarized by dividing them into two types, 
those intended to predict temperature rise in soft tissue, and those for bone. In the first 
case, for soft-tissue heating, the formulae have the form TI = Af W, where/is the acoustic 
frequency, and W is the acoustic power. The constant A takes defined values for the 
specific scan modes and geometry (for example it is 1/210 for B-mode scanning). For 
exposure of bone, the formulae are of the form TI = B W, where the constant B again takes 
defined values. TI for bone does not depend on the acoustic frequency. 



TI formulae also differ as to where the temperature rise is estimated. For stationary 
beams, for example for M-mode or spectral pulsed Doppler, greatest temperatures are 

138 



Safety standards and regulations: the manufacturers' responsibilities 10 



reached deeper within the tissue (for example see Figure 10.1b). On the other hand it is 
assumed that when beams are scanned the greatest heating is at the surface (Figure 10.1a). 
A thorough discussion of the rationale behind the formulae is given by Abbott (1999). 

The ODS also prescribes the conditions for which the safety indices are to be displayed. In 
general, only one index value needs to be displayed, an index need not be displayed if its 
value is less than 0.4, and need not be displayed at all if it can never exceed 1.0 under any 
condition. In fact many manufacturers display values more fully than is strictly required 
by the ODS, both by showing values below 0.4 and by not restricting the display to a 
single index value. 

TI is based upon calculations of temperature rise in tissue, but must not be considered as 
an accurate estimate of temperature rise. For example, the assumption that temperature 
rise can be predicted simply from an analysis of the absorption of ultrasound in tissue has 
been shown to be false for positions close to the transducer. This is because the transducer 
itself heats slightly, so warming the adjacent tissue by thermal conduction. Another error 
relates to the way in which the heating contributions are added together when several 



Figure 10.1. Simple models used for the calculation of (a) soft-tissue thermal index (TIS) and 
the (b) bone-at-focus thermal index, TIB. Soft tissue is modelled as a homogeneous material 
with an attenuation coefficient of 0.3dBcm _1 MHz -1 . The temperature variation along the 
axis reaches a maximum close to the transducer for soft-tissue exposure (TIS) and reaches a 
maximum at the focus for the bone-at-focus exposure (TIB). 



Maximum heating 
is assumed to be 
at the surface 
when imaging, 
and at depth 
for Doppler and 
M-mode 

It is not 
always 
required to 
display either 
Ml or TI 



Some 
problems 
remain 
with the 
safety index 
formulations 



(a) 




TIS 



(b) 




TIB 



139 



10 Safety standards and regulations: the manufacturers' responsibilities 



diagnostic modes are used together. An additional difficulty arises from the incorrect 
assumption that in situ exposure can be calculated from measurements in water using 
simple linear assumptions, a problem that can cause underestimation of the MI. 



10.4. European Medical Devices Directive 



CE marking 
for medical 
equipment 
requires 
conformance 
with the EU 
MDD 



The European Communities Medical Devices Directive or MDD (European Communities, 
1993) is adopted into the national legal framework of many European member states, 
including the UK. Each state appoints a Competent Authority. For the UK this is the 
Medicines and Healthcare Products Regulatory Agency (MHRA). The Agency designates 
Notified Bodies whose function is to make assessments of new products under the MDD 
procedures. Manufacturers can use any appropriate Notified Body within the European 
Union. A CE mark on a device means that the device satisfies the requirements essential 
for it to be fit for its intended purpose. All medical devices (except custom-made and 
devices intended for clinical investigations), whether used in private or public hospitals 
and nursing homes, or sold in retail outlets, must carry the CE marking. 



Classification 
of equipment 
depends on 
the level of 
risk 



Annex IX of the MDD describes three general classes of equipment and a set of rules for 
establishing the class of any particular type of equipment. Almost all diagnostic ultrasonic 
devices lie in Class Ha, because they are "active devices intended for diagnosis". Class 
lib is specifically intended for monitoring vital physiological parameters, where the 
nature of the variations is such that they could result in immediate danger to the patient. 
Some Doppler devices used during surgery may be deemed to fall into this category. 
Intravascular ultrasound transducers are usually designated as Class III since they are 
invasive devices. Initially the manufacturer determines the class of his products, and 
selects an appropriate Notified Body to carry out the conformity assessment procedure. 
The assessment in all categories requires appropriate audits of manufacturer's production 
quality assurance systems. In addition, there is a provision for examination and testing of 
products or batches, although this is not mandatory, even for Class III products, provided 
that the design dossier is submitted. It is probable that most medical ultrasound devices 
gain a CE mark without practical independent assessment, being based substantially on 
inspection of required documentation. Manufacturers are also required to maintain a 
post-market surveillance system and report certain types of incidents to a Competent 
Authority. 



Manufacturers 
must 
demonstrate 
both output 
safety and 
measurement 
accuracy 



The MDD does not give any guidance specific to ultrasound emissions. The general 
statements of Clause 11 (on protection against radiation) and of Clause 12 (on equipment 
with an energy source) may be taken to apply, however. These require that "Devices 
shall be designed and manufactured in such a way that exposure to patients . . . shall be 
reduced as far as possible compatible with the intended purpose, whilst not restricting 
the application of appropriate specified levels for therapeutic or diagnostic purposes". 
Devices may emit hazardous levels of radiation if necessary for a specific medical purpose, 
but it must be possible to control them. Where radiation is "potentially hazardous", 
devices must be "fitted, where practicable, with visual displays and/or warnings of such 
emissions". Furthermore, "accessible parts of devices ... must not attain potentially 
dangerous temperatures under normal use". The normal method for a manufacturer to 

140 



Safety standards and regulations: the manufacturers' responsibilities 10 

demonstrate that they have complied with these requirements is to follow procedures 
laid down in international standards. 

References 

Abbott JG. 1999. Rationale and derivation of MI and TI— a review. Ultrasound Med Biol, 
25,431-441. 

American Institute of Ultrasound in Medicine/National Electrical Manufacturers 
Association. 1998. Standard for Real-time Display of Thermal and Mechanical Acoustic 
Output Indices on Diagnostic Ultrasound Equipment. 2nd Edn. Rockville, MD: AIUM. 
European Communities. 1993. Council directive 93/42/EEC, of 14 June 1993, concerning 
medical devices. Off J Eur Communities, 36, L169. 

Food and Drug Administration: US Department of Health and Human Services. 
1997. Information for Manufacturers Seeking Marketing Clearance of Diagnostic 
Ultrasound Systems and Transducers. www.fda.gov/downloads/MedicalDevices/ 
DeviceRegulationandGuidance/GuidanceDocuments/UCM07091 1 .pdf . Rockville, 
MD: Center for Devices and Radiological Health. 

IEC. 2001. IEC60601 Part 2-37: Medical Electrical Equipment: Particular Requirements for 

the Safety of Ultrasound Diagnostic and Monitoring Equipment 2001 & Amendment 1 

2005. Geneva, Switzerland: International Electrotechnical Commission. 

IEC. 2005. IEC60601 Part 1: Medical Electrical Equipment: General Requirements for 

Safety and Essential Performance. Geneva, Switzerland: International Electrotechnical 

Commission. 

IEC. 2006. IEC62359: Ultrasonics - Field Characterization - Test Methods for the 
Determination of Thermal and Mechanical Indices Related to Medical Diagnostic 
Ultrasound Fields. Geneva, Switzerland: International Electrotechnical Commission. 



141 



The Safe Use of Ultrasound in Medical Diagnosis 



Chapter 1 1 

Guidelines and recommendations 
for the safe use of diagnostic 
ultrasound: the user's 
responsibilities 

Gail ter Haar 

Institute of Cancer Research, Sutton, UK 



Summary 

• The responsibility for risk assessment has migrated from regulatory authorities to 
the user. 

• There is a strong need for continuing education to ensure that appropriate risk/benefit 
assessments are made. 

• Major national ultrasound societies have formulated guidelines for the safe use of 
ultrasound in medicine. 

11.1 Introduction 

The evolution of ultrasonographic equipment has led to the development of powerful 
diagnostic devices that are capable of substantial acoustic output. Coincidentally, 
the method of ensuring safe use of diagnostic ultrasound in medicine has undergone 
significant change towards a system of self-regulation. This change originated in the 
USA. This shift in responsibility for risk assessment from a regulatory authority to the 
user has created an urgent need for responsible attitudes to safety issues. To encourage 
this approach it is encumbent on authorities, ultrasound societies and expert groups 
to provide relevant information about the risk of producing biological effects during 
ultrasonographic procedures. There is a strong need for continuing education to ensure 
that appropriate risk/benefit assessments are made, based on an understanding of the 
probability of biological effects occurring with each type of ultrasound procedure. 
Some types of examinations involve a greater risk of adverse effect than do others. 
The probability of causing some of these bio-effects can be increased by the presence 
of gas bubbles such as are used for echo-contrast agents. Also, the sensitivity of the 

142 



Guidelines and recommendations 11 



tissue target has a significant impact on the type of bio-effect and on its consequences 
for human health. 

The relaxation of intensity limits for pre-market approval by the United States Food 
and Drug Administration (FDA, 1985, 1993) allows a substantially increased intensity 
of ultrasound to be delivered to the foetus than was previously allowed. Studies with 
laboratory animals have shown that significant biological effects can be produced from 
exposures from modern diagnostic equipment operating at maximum output conditions. 
These issues have been addressed by the World Federation for Ultrasound in Medicine 
and Biology (WFUMB) in published conclusions and recommendations that represent 
international consensus on safe use of ultrasound in medicine. Major national ultrasound 
societies have formulated guidelines for the safe use of ultrasound in medicine. This 
chapter describes relevant published safety guidelines issued by national and international 
organizations concerned with the safe and effective application of ultrasound in medicine. 



An important question that relates to the safe use of ultrasound in medicine is whether 
it is better to limit the acoustic output by law or to rely on optimum patient exposure 
and appropriate risk/benefit assessment based on informed use. The latter option is 
encouraged by the FDA through its acceptance of equipment output display using the so- 
called "Track 3" option for pre-market equipment approval (see Chapter 10). However, 
this option requires effort on the part of the user, authorities and ultrasound organizations 
to disseminate relevant information about potential risk. 



Appropriate use 
of safety related 
"output display" 
requires user 
education and 
dissemination 
of relevant 
information about 
potential risk 



The change from a regulated maximum intensity limit to control by personal judgement 
of the risk/benefit ratio for each type of examination allows the clinician or ultrasound 
technologist access to, and control of, potentially substantially higher acoustic outputs 
than before. It is appropriate that clinicians take the responsibility of risk assessment 
armed with information on bio-effects provided by the scientific community, and on 
equipment output conditions provided by the manufacturer. This increasingly places 
the responsibility on the end-user to maximize the benefit of ultrasound examinations, 
whilst minimizing the risk. Clearly, in order to make valid judgements radiologists and 
sonographers must be educated about safety issues. Attitudes that assume inherent 
safety under all conditions, simply because the equipment is commercially available, are 
inappropriate. Misplaced assumptions that the highest acoustic power always gives the 
best diagnostic information need to be replaced by an awareness of the relative risks for 
each application. Ultrasound technologists who have been trained in safety issues are 
best able to interpret information in the form of an output display, and it is the purpose 
of this chapter to draw attention to some important and relevant issues. It is essential that 
users of ultrasound equipment have an understanding of bio-effect mechanisms and are 
cognizant of the potential risks associated with different modes of operation. The risk/ 
benefit changes significantly according to the medical reason for undertaking each type 
of ultrasonographic examination. For example, a current topic of debate is whether or not 
there is risk, or benefit, associated with routine use of pulsed Doppler (PD) ultrasound 
during the first trimester in uncomplicated pregnancy. While there is some uncertainty 
about the risk/benefit issues, it is well known that the developing embryo is particularly 
sensitive to damage by physical insult. It should also be remembered that both the 

143 



The introduction 
of displayed safety 
"indices" has 
resulted in higher 
available acoustic 
outputs 



11 Guidelines and recommendations 



benefit and risk depend largely on the skill and competence of the person performing 
the examination. 

There is a large data base of scientific papers and review articles on biological effects 
and safety of ultrasound. The WFUMB has published proceedings of symposia offering 
conclusions and recommendations on the safe use of ultrasound in medicine. The 
WFUMB Safety Committee has also published relevant scientific review papers (Barnett 
et al., 1994, 1997, 2010). Most national ultrasound societies have a safety committee or 
advisory group whose role is to provide information on the risk of bio-effects. Some of 
these expert groups also issue safety guidelines to promote responsible use of diagnostic 
ultrasound. Information on bio-effects and safety is disseminated through regular 
scientific presentations at medical ultrasound conferences. However, the low audience 
attendance at bio-effects sessions at clinical conferences suggests that the concept of 
voluntary attention to safety issues may not be appropriate or effective. The situation is 
made more difficult by the unknown number of clinicians and sonographers who practice 
outside tertiary centres and who do not have any affiliation with ultrasound societies. The 
purpose of this chapter is to draw attention to existing safety guidelines and to provide 
some background information on the current status of knowledge on biophysical aspects 
of ultrasound interaction. 



Complacency 
about safety 
should be 
avoided. Each 
new technique 
needs separate 
evaluation 



It is easy to become complacent in the knowledge that medical ultrasonography has 
enjoyed widespread use as a safe and effective diagnostic clinical tool, but it must be 
remembered that the fact that there is no evidence for harm does not provide evidence 
of absence of potential to do harm. However, it should be realized that each new 
technological development introduces new biophysical conditions that require evaluation 
with regard to safety; for example, echo-contrast agents enhance cavitation and harmonic 
generation in tissue. Improvements in resolution, grey-scale definition and image quality 
are particularly important in obstetrics. Image quality has been further enhanced with 
the advent of endovaginal examinations that allow closer access to anatomical structures 
and visualization of the embryo or foetus without suffering the beam-interference effects, 
or ultrasound attenuation, caused by overlying abdominal skin, fat and musculature. 
PD spectral flow analysis and Doppler colour flow imaging (CFI) techniques offer the 
potential to increase diagnostic effectiveness and may be attractive for applications in 
early pregnancy. Meanwhile, there is a developing trend towards prolonged daily use of 
transcranial PD spectral flow measurement in premature infants. 



Non-linear harmonic imaging introduces another interesting variable into the potential 
bio-effects equation. Higher-frequency harmonics are generated in tissue by the use of 
high acoustic pressure amplitudes. In addition the introduction into the ultrasound field 
of echo-contrast gas bodies can generate harmonics locally. By these processes the non- 
linear effect is enhanced and the potential for biological effects may be also increased (see 
Chapters 2 and 6). 



Major ultrasound societies and organizations have paid attention to the safety of 
diagnostic ultrasound and progress has been achieved towards identifying international 
consensus on important issues. In recent years, knowledge of the biophysical effects of 

144 



Guidelines and recommendations 11 



ultrasound has become a fundamental consideration in the process of setting international 
standards for safety of ultrasound in medicine, an important function of the International 
Electrotechnical Committee (IEC, 1992; see Chapter 10). 

It is entirely appropriate to recognize that the scientific community does not have all the 
answers to basic questions, and that fresh questions are generated as new applications 
develop. This uncertainty may require the introduction of wide safety margins in 
international standards in order to ensure that risk is minimized. 



The regulatory arrangements set up by the FDA in the USA, and its use of the AIUM/ Safety indices 

NEMA (1992) " Output Display Standard" (ODS) are discussed in Chapter 10. However, t a ^ 0 a ^ c ° unt 

there are some important issues that need to be understood when considering the 0 f the exposure 

AIUM/NEMA output labelling scheme. The duration of exposure is not included in the 

index of risk, and this remains an important consideration, particularly for thermally 

mediated biological effects where the damage threshold is defined by a combination 

of temperature increase and time for which it is maintained (see Chapter 3). The 

higher the temperature increase, the shorter the duration required to produce adverse 

effects. This issue has been examined in detail by international panels of experts 

during symposia sponsored by the WFUMB. Following its safety symposium in 1996, 

the WFUMB published a recommendation (WFUMB, 1998) that "... safety guidelines 

should include an appropriate duration factor". This aspect of safety issues has been 

recognized by the IEC in its deliberations towards developing international safety 

standards. 



Another potential limitation of the thermal index (TI) is that it does not take account of 
patient temperature. This may be an important consideration for obstetrics. A relevant 
recommendation of the WFUMB (1998) states that "Care should be taken to avoid unnecessary 
additional embryonic and foetal risk from ultrasound examinations of febrile patients". 



TI takes no 
account 
of patient 
temperature 



In comparison with thermal mechanisms, gas-body effects from diagnostic ultrasound 
can occur almost instantaneously when the acoustic pressure exceeds a certain threshold 
value. The ODS does not give any indication of exposure duration or the effects on bio- 
effects thresholds of such factors as the presence of gas bodies/contrast agents, patient 
temperature or non-linear propagation. 



The European Committee for Ultrasound Radiation Safety (ECURS) has published an 
informative tutorial paper on the thermal and mechanical indices (EFSUMB, 1996a,b) 
which draws attention to some of these problems. As an example, it suggests that the 
models used for deriving "reasonable worst-case" exposures for the indices may not be 
adequate to describe first-trimester scanning through a full bladder where the attenuating 
effect of overlying tissue is small. The purpose of the mechanical index (MI) is to predict 
the likelihood of cavitation type of bio-effects where the peak pressure amplitude is a 
critical parameter. The intended use is for the display of MI to be included in B-mode 
imaging. However, recent surveys (Duck and Henderson, 1998; Henderson et al., 1997; 
Martin, 2010) show minimal differences in peak pressure values for different modes of 
operation (see Chapter 3). 

145 



11 Guidelines and recommendations 



The importance of an MI, or similar indicator of risk of mechanical damage, is emphasized 
by recent research findings. Although there is no direct evidence of adverse effects 
in humans (no scientifically controlled studies with humans exist) from mechanical 
bio-effects, the scientific literature contains a considerable amount of information on both 
in vitro and in vivo effects in lower animals and mammals. The critical acoustic parameter 
is the in situ rarefaction pressure. 

Surveys of acoustic output measurements reported for equipment in clinical use in the 
UK (Henderson et al, 1997; Whittingham, 2000) have indicated that the peak rarefactional 
(p r ) pressure amplitudes for Doppler systems range between 0.6 MPa and 5.5 MPa. A more 
recent survey of values declared by manufacturers puts this range as 2.1-6.7 MPa (Martin, 
2010). For imaging systems, the measured range was 0.5-4.6 MPa, and the declared range 
was 2.3-6.4 MPa. The observation of haemorrhage in the mouse lung following exposure 
to pulsed ultrasound (Child et al., 1990) at 1 MPa has demonstrated that these rarefactional 
pressure amplitudes are sufficient to lead to adverse biological effects in mammalian 
tissues. It is not certain to what extent these effects may occur in humans or their clinical 
significance (WFUMB, 1998; see Chapter 5). 

11.2 International guidelines 

11.2.1 World Federation for Ultrasound in Medicine 
and Biology 

Since 1985, the WFUMB has sponsored a number of symposia on safety and standardization 
of ultrasound in medicine that have addressed topics related to known mechanisms for 
producing biological effects. These recommendations remain valid today, and are the 
basis for current recommendations. During these symposia the available scientific data 
base has been critically examined, conclusions drawn and the proceedings subject to 
widespread international review. As a result the WFUMB has published policy statements 
and a number of conclusions and recommendations, endorsed as internationally 
acceptable guidelines for the safe use of diagnostic ultrasound (WFUMB, 1992, 1998). 
These guidelines are also still valid today. 

11.2.1.1 WFUMB policy statement on safety of diagnostic ultrasound 

B-mode imaging 

Known diagnostic ultrasound equipment as used today for simple B-mode imaging operates at 
acoustic outputs that are not capable of producing harmful temperature rises. Its use in medicine 
is therefore not contraindicated on thermal grounds. This includes endoscopic, transvaginal and 
transcutaneous applications. 

Doppler 

It has been demonstrated in experiments with unperfused tissue that some Doppler diagnostic 
equipment has the potential to produce biologically significant temperature rises, specifically at 
bone/soft tissue interfaces. The effects of elevated temperatures may be minimized by keeping the 
time for which the beam passes through any one point in tissue as short as possible. Where output 



146 



Guidelines and recommendations 11 



power can be controlled, the lowest available power level consistent with obtaining the desired 
diagnostic information should be used. Although the data on humans are sparse, it is clear from 
animal studies that exposures resulting in temperatures less than 38.5 °C can be used without 
reservation on thermal grounds. This includes obstetric applications. 

Transducer heating 

A substantial source of heating may be the transducer itself. Tissue heating from this source is WFUMB 

localized to the volume in contact with the transducer. policy 

statement 

The findings of the WFUMB symposia on the safety of ultrasound in medicine are 
published in the scientifi c literature. These reports (WFUMB, 1992, 1998) contain 
comprehensive lists of conclusions and recommendations on a wide range of biological 
effects. Some of the clinically relevant conclusions and recommendations on the safe use 
of ultrasound in medicine are given below. 



11.2.1.2 WFUMB conclusions 

Thermal effects 

Developmental abnormalities have been observed in animals when the embryonic or foetal 
temperature is increased by 2 °C or more above their normal body temperature for extended duration. 



Biologically significant temperature increases have been measured, at or near bone/soft tissue 
interfaces, during exposure to conditions similar to those used in Doppler diagnostic equipment. 
The effects of elevated temperatures may be minimized by keeping the time for which the beam 
passes through any one point in tissue as short as possible. 

Non-thermal effects 

Capillary bleeding has been observed in the lung after exposure of neonatal, young and adult mice, 
swine and adult rats, rabbits and monkeys to diagnostically relevant, pulsed ultrasound. Thresholds 
for capillary bleeding in adult mice and neonatal and young swine are of the order ofl MPa at 2 MHz, 
which is within the range of output values of commercially available diagnostic ultrasound systems. 

In the air-filled mammalian lung, bleeding from alveolar capillaries has been induced experimentally by WFU MB 
ultrasound at diagnostic exposure levels. This effect has not been observed in the fluid-filled mammalian conclusii 
foetal lung. There is no direct evidence to date as to whether or not this effect can occur in humans. 



11.2.1.3 WFUMB recommendations 

Thermal effects 

A diagnostic exposure that produces a maximum temperature rise of no more than 1.5 °C above 
normal physiological levels (37 °C) may be used clinically without reservation on thermal grounds. 
A diagnostic exposure that elevates embryonic and foetal in situ temperature 41 °C 
(4 °C above normal temperature) for 5 min or more should be considered potentially hazardous. 



The risk of adverse effects of heating is increased with the duration of exposure. Thus, safety 
guidelines should include an appropriate duration factor. 

147 



11 Guidelines and recommendations 



Care should be taken to avoid unnecessary additional embryonic and foetal risk from ultrasound 
examinations of febrile patients. 



Non-thermal effects 

Cavitation: It has been shown experimentally that acoustic cavitation can alter mammalian 
tissue. The possible occurrence of cavitation, either inertia! or non-inertial, should be considered 
in assessing the safety of diagnostic ultrasound and of other forms of medical ultrasound. 

Lung capillary bleeding: Currently available animal data indicate that it is prudent to reduce 
ultrasound exposure of human postnatal lung to the minimum necessary to obtain the required 
diagnostic information. 

A risk benefit analysis should be performed if the anticipated acoustic pressure amplitude at the 
surface of the postnatal lung exceeds IMPa. 



Contrast agents: Gas bodies introduced by a contrast agent increase the probability of cavitation. 
A physician should take this into account when considering the benefit/risk ratio of an 
examination. 

WFUMB 

recommendations B-mode imaging: When tissue/gas interfaces or contrast agents are not present, the use ofB-mode 
imaging need not be withheld because of concern for ultrasound safety. This statement also 
applies to endoscopic, transvaginal and transcutaneous applications. When tissue/gas interfaces 
or contrast agents are present, ultrasound exposure levels and duration should be reduced to the 
minimum necessary to obtain the required diagnostic information. 

As mentioned above, the main area of concern in obstetric ultrasound scanning lies with 
the use of Doppler in the first trimester. Here several organizations (WFUMB, ISUOG, 
EFSUMB/AIUM) have agreed a single statement (WFUMB, 2011). This is: 



Statement on the Safe Use of Doppler Ultrasound During 1 1-14 week scans (or earlier in pregnancy) 



WFUMB/EFSUMB/ 
AIUM/ISUOG 
Doppler in the 
first trimester 



1. PD (spectral, power and CFI) ultrasound should not be used routinely. 

2. PD ultrasound may be used for clinical indications such as to refine risks for trisomies. 

3. When performing Doppler ultrasound, the displayed TI should be less than or equal to 1.0 and 
exposure time should be kept as short as possible (usually no longer than 5-10 min) and not 
exceed 60 min. 

4. When using Doppler ultrasound for research, teaching and training purposes, the displayed 
TI should be less than or equal to 1.0 and exposure time should be kept as short as possible 
(usually no longer than 5-10 min) and not exceed 60 min. Informed consent should be 
obtained. 

5. In educational settings, discussion of first trimester pulsed or colour Doppler should be 
accompanied by information on safety and bio-effects (e.g. TI, exposure times and how to 
reduce the output power). 

6. When scanning maternal uterine arteries in the first trimester, there are unlikely to be any 
foetal safety implications as long as the embryo/foetus lies outside the Doppler ultrasound 
beam. 



148 



Guidelines and recommendations 11 



11.2.2 Other international safety guidelines 

The other professional Ultrasound societies also issue statements and guidelines relevant 
to ultrasound safety. Each has its own form of clinical safety statement, but the message 
is consistent. It is important to understand that these statements on clinical ultrasound 
safety are not regulatory. They provide essential information and serve a useful role in 
advising users of the potential risk from certain diagnostic ultrasound procedures. The 
most recent such statement is from EFSUMB, and is reproduced here. The interested 
reader is referred to the websites of AIUM, ISUOG and ASUM (Australian Society for 
Ultrasound in Medicine) for further information. 

11.2.2.1 EFSUMB Clinical Safety Statement for Diagnostic Ultrasound (2011) 

Diagnostic ultrasound has been widely used in clinical medicine for many years with no proven 
deleterious effects. However, if used imprudently, diagnostic ultrasound is capable of producing 
harmful effects. The range of clinical applications is becoming wider, the number of patients 
undergoing ultrasound examinations is increasing and new techniques with higher acoustic 
output levels are being introduced. It is therefore essential to maintain vigilance to ensure the 
continued safe use of ultrasound. 

Ultrasound examinations should only be performed by competent personnel who are trained 
and updated in safety matters. It is also important that ultrasound devices are appropriately 
maintained. 

Ultrasound produces heating, pressure changes and mechanical disturbances in tissue. Diagnostic 
levels of ultrasound can produce temperature rises that are hazardous to sensitive organs and the 
embryo/foetus. Biological effects of non-thermal origin have been reported in animals but, to date, no 
such effects have been demonstrated in humans, except when a microbubble contrast agent is present. 

The TI is an on-screen guide to the user of the potential for tissue heating. The MI is an 
on-screen guide of the likelihood and magnitude of non-thermal effects. Users should regularly 
check both indices while scanning and should adjust the machine controls to keep them as low 
as reasonably achievable (ALARA principle) without compromising the diagnostic value of the 
examination. Where low values cannot be achieved, examination times should kept as short as 
possible. Guidelines issued by several ultrasound societies are available. 

Some modes are more likely than others to produce significant acoustic outputs and, when using 
these modes, particular care should be taken to regularly check the TI and MI indices. Spectral 
pulse wave Doppler and Doppler imaging modes (CFI and power Doppler imaging) in particular 
can produce more tissue heating and hence higher TI values, as can B-mode techniques involving 
coded transmissions. Tissue harmonic imaging mode can sometimes involve higher MI values. 3D 
(three dimensional) imaging does not introduce any additional safety considerations, particularly 
if there are significant pauses during scanning to study or manipulate the reconstructed images. 
However, 4D scanning (real-time 3D) involves continuous exposure and users should guard 
against the temptation to prolong examination times unduly in an effort to improve the recorded 
image sequence beyond that which is necessary for diagnostic purposes. 

149 



11 Guidelines and recommendations 



Ultrasound exposure during pregnancy 

The embryo/foetus in early pregnancy is known to be particularly sensitive. In view of this 
and the fact that there is very little information currently available regarding possible subtle 
biological effects of diagnostic levels of ultrasound on the developing human embryo or foetus, 
care should be taken to limit the exposure time and the thermal and mechanical indices to the 
minimum commensurate with an acceptable clinical assessment. 

Temperature rises are likely to be greatest at bone surfaces and adjacent soft tissues. With 
increasing mineralization of foetal bones, the possibility of heating sensitive tissues such as 
brain and spinal cord increases. Extra vigilance is advised when scanning such critical foetal 
structures, at any stage in pregnancy. Based on scientific evidence of ultrasound-induced 
biological effects to date, there is no reason to withhold diagnostic scanning during pregnancy, 
provided it is medically indicated and is used prudently by fully trained operators. This includes 
routine scanning of pregnant women. However, Doppler ultrasound examinations should not be 
used routinely in the first trimester of pregnancy. 

The power levels used for foetal heart rate monitoring (cardiotocography) are sufficiently low that 
the use of this modality is not contraindicated on safety grounds, even when it is to be used for 
extended periods. 

Safety considerations for other sensitive organs 

Particular care should be taken to reduce the risk of thermal and non-thermal effects 
during investigations of the eye and when carrying out neonatal cardiac and cranial 
investigations. 

Ultrasound contrast agents 

These usually take the form of stable gas-filled microbubbles, which can potentially produce 
cavitation or microstreaming, the risk of which increases with MI value. Data from small animal 
models suggest that microvascular damage or rupture is possible. Caution should be considered 
for the use of ultrasound contrast agents (UCA) in tissues where damage to microvasculature 
could have serious clinical implications, such as in the brain, the eye and the neonate. As in all 
diagnostic ultrasound procedures, the MI and TI values should be continually checked and kept as 
low as possible. It is possible to induce premature ventricular contractions in contrast-enhanced 
echocardiography when using high MI and end-systolic triggering. Users should take appropriate 
precautions in these circumstances and avoid cardiac examinations in patients with recent acute 
coronary syndrome or clinically unstable ischaemic heart disease. The use of contrast agents should 
be avoided 24 h prior to extra-corporeal shock wave therapy. 

In conclusions and recommendations on transvaginal sonography EFSUMB advises that 
(EFSUMB 1995): 

"The absence of long-term, large scale, follow-up studies following first-trimester 
ultrasound exposures means that care is required in the application of transvaginal 



150 



Guidelines and recommendations 11 



ultrasonography in early pregnancy. It should only be performed for pure medical 
reasons that are to the benefit of the mother and/or the embryo" . 

11.2.2.2 British Medical Ultrasound Society 

The British Medical Ultrasound Society (BMUS) is the only organization to have 
published guidelines on the safe use of ultrasound, with explicit recommendations 
related to the displayed safety indices (BMUS, 2010). This information is shown in Tables 
11.1 and 11.2. Figure 11.1 shows a graphical representation providing easy reference 
for the recommended scanning time at any given TI for obstetric scans. Similar figures 
are available for neonatal transcranial and spinal scans, general neonatal and cardiac 
scans, and adult transcranial, peripheral vascular and general abdominal scanning. 
For obstetrics, TIs of 0.7 can be used without restriction, and the recommendation 
is that those greater than 3 should be avoided. The time limits were reached using 
the thermal guidelines produced by WFUMB (1998), rounding the times down where 
appropriate to ensure continued safety. This is discussed in more detail in Chapter 3. 
In discussing the recommended time restrictions, the Guidelines state that: 

"The operator should aim to stay within BMUS recommended scan times. If there is a 
clinical need to exceed these recommended times, the ALARA principle should still be 
followed. When overall times longer than those recommended here are essential, the 
probe should be removed from the patient whenever possible, to minimize exposure." 

11.3 Souvenir scanning 

All the societies mentioned above have issued statements about the use or the production 
of souvenir scans (also known as keepsake, or bonding scans). The message of these 
statements is that this practice cannot be recommended, giving safety grounds as the 
basis for this (WFUMB, AIUM, ISUOG, EFSUMB, BMUS). At first sight this may seem to 
contradict the " clinical" safety statements from these same organizations in which routine 
scans during pregnancy are said to be safe. An ultrasound scan conducted in a souvenir 
scanning centre is not inherently more harmful than the same scan conducted for clinical 
reasons if carried out by a qualified practitioner. The difference lies in the perceived 
benefit obtained compared to any potential risk. A "routine" obstetric ultrasound scan 
is conducted with the expectation that it will inform the management of the pregnancy 
beneficially, whereas a souvenir scan is carried out solely for "recreational" purposes. 
A compromise is reached by conceding that providing a "souvenir" image at the end 
of a clinically indicated scan does not add significantly to any potential risk, and may 
dissuade the pregnant mother from resorting to a high street "boutique" with unknown 
skills and qualifications to obtain such a scan (Brezinka, 2010; Phillips et ah, 2010). 

The EFSUMB, ISUOG and WFUMB statements are given below. The EFSUMB statement 
has been endorsed as BMUS policy. 



151 



11 Guidelines and recommendations 



Table 11.1. Recommended exposure time and index values for obstetric and neonatal ultrasound. 



Application 


Values to 

monitor 

(A) 


TI value 


Ml value 




0- 
0.7 


0.7-3.0 


>3.0 


0- 
0.3 


>0.3 


>0.7 


Obstetrics up 
10 weeks after 
LMP (and 
gynaecology 
when 

pregnancy is 
possible) 


TIS and 
MI 


■/ 


(B) restrict time to 

0. 7 < TIS < 1.0: 60 min 

1. u i io l.d. ju min 
1.5 < TIS < 2.0: 15 min 
2.0 < TIS < 2.5: 4 min 
2.5 < TIS < 3.0: Imin 


Scanning of an 
embryo or 
foetus is not 
recommended, 
however briefly 


■/ 


■/ 


(E) risk of 

cavitation 

with 

contrast 

agents 


Obstetrics 
more than 10 
weeks after 
LMP 


TIB and 
MI 


■/ 


(B) restrict time to 
0.7 < TIB < 1.0: 60 min 
1.0 < TIB < 1.5: 30 min 
1.5 < TIB < 2.0: 15 min 
2.0 < TIB < 2.5: 4 min 
2.5 < TIB < 3.0: 1 min 


Scanning of an 
embryo or 
foetus is not 
recommended, 
however briefly 


■/ 


■/ 


(E) risk of 

cavitation 

with 

contrast 

agents 


Neonatal— 
transcranial 
and spinal 


TIC and 
MI 




(B) restrict time to 
0.7 < TIC < 1.0: 60 min 
1.0 < TIC < 1.5: 30 min 
1.5 < TIC < 2.0: 15 min 
2.0 < TIC < 2.5: 4 min 
2.5 < TIC < 3.0: Imin 


Scanning of an 
embryo or 
foetus is not 
recommended, 
however briefly 






(E) risk of 

cavitation 

with 

contrast 

agents 


Neonatal— 
general 
and cardiac 
imaging 


TIB and 
MI recom- 
mended 




fC) restrict time to 
1.0 < TIB < 1.5: 120 min 3.0 < TIB < 4.0 : 1 min 
1.5 < TIBS 2.0: 60 min 4.0 < TIB < 5.0 : 15 s 
2.0 < TIB < 2.5: 15 min 5.0 < TIB < 6.0 : 5 s 
2.5 < TIB < 3.0: 4 min TIB > 6: not 
recommended. 




(D) pos- 
sibility 
of minor 
damage 
to lung or 
intestine. 
Minimize 
exposure 
time. 


(E) risk of 

cavitation 

with 

contrast 

agents 


Foetal 

Doppler heart 
monitoring 


TI or MI 
are not 
usually 
available 
for 

dedicated 
foetal 
heart 
monitors 


The power levels used by dedicated foetal heart monitors are sufficiently low 

that the use of this modality is not contraindicated, 

on safety grounds, even when it is to be used for extended periods 



S : There is no known reason to restrict scanning times in this region. 

A: Many scanners allow MI and one of the TI values to be displayed simultaneously: the most appropriate 
TI value depends on the clinical application. 

B: TI > 0.7— the overall exposure time (including pauses) of an embryo or foetus or of the neonatal central 
nervous system should be restricted. 

C: TI> 1.0— the overall exposure time (including pauses) of other parts of the neonate should be restricted. 
D: MI > 0.3— there is a possibility of minor damage to neonatal lung or intestine. If such exposure is 
necessary, try to reduce the exposure time as much as possible. 

E: MI > 0.7— there is a risk of cavitation if an ultrasound contrast agent containing gas microspheres is being used. 
There is a theoretical risk of cavitation without the presence of ultrasound contrast agents. The risk increases with 
MI values above this threshold. 



152 



Guidelines and recommendations 11 



Table 1 1 .2. Recommended exposure time and index values for non-obstetric and non- neonatal 
ultrasound. 



Application 


Values to monitor 
(A) 


TI value 


Ml value 




0-1.0 


>1.0 


0-0.3 


>0.7 


General 
abdominal 

Peripheral 
vascular 

Unlisted 
applications 


Usually TIB and MI 

[use TIC and MI if 
bone closer than 
1 cm; TIS and MI 
only if bone does not 
come into the image] 


■/ 


(B) restrict time to 
1.0 < TIB < 1.5: 120 min 
1.5 < TIB < 2.0: 60 min 
2'0<TIB<2'5: 15 min 
2.5 < TIB < 3.0: 4 min 
3.0 < TIB < 4.0:1 min 
4.0 < TIB < 5.0: 15 s 
5.0 < TIB < 6.0: 5 s 
TIB>6: not 
recommended 


■/ 


(C) risk of cavitation 
with contrast agents 


Eye 


TIS and MI 
recommended 


■/ 


Scanning of the eye is 
not recommended 


V 


(C) risk of cavitation 
with contrast agents 








(B) restrict time to 






Adult 
transcranial 
(imaging and 
stand-alone) 
(D) 


TIC and MI 




0.7 < TICS 1.0: 60 min 
1.0 < TIC < 1.5: 30 min 
1.5 < TIC < 2.0: 15 min 
2.0 < TIC < 2.5: 4 min 
2.5 < TIC < 3.0: 1 min 
TIC>3: not 
recommended 




(C) risk of cavitation 
with contrast agents 


Peripheral 
pulse 

monitoring 


TI or MI are not 
usually available 
for dedicated 
peripherial pulse 
monitoring 


The output from CW Doppler devices intended for monitoring 
peripheral pulses is sufficiently low that their use is not 
contraindicated, on safety grounds 



/ : There is no known reason to restrict scanning times in this region. 

A: Many scanners allow MI and one of the TI values to be displayed simultaneously: the most appropriate 
TI value depends on the clinical application. 

B: TI> 1.0— the overall exposure time (including pauses) should be restricted. 

C: MI > 0.7— there is a risk of cavitation if an ultrasound contrast agent containing gas microspheres is being 
used. There is a theoretical risk of cavitation without the presence of ultrasound contrast agents. The risk 
increases with MI values above this threshold. 

D: Transcranial ultrasound investigations may require higher acoustic output or longer monitoring times 
than other applications. When times longer than those recommended here are required, it is recommended 
that monitoring is paused regularly to minimize exposure. 



153 



11 Guidelines and recommendations 



OBSTETRIC SCANNING 

THERMAL INDEX 

0 0.5 1.0 1.5 2.0 2.5 3.0 

I . I I I L 



0.7 












^ w 

RECOMMENDED 
RANGE 
PROVIDED 
ADEQUATE IMAGES 
CAN BE OBTAINED 
(especially in 1st trimester) 

Unlimited time 
Observe ALARA 


<60 
mins 


* * 

<30 
mins 


* * 

<15 
mins 


* * 

<4 
mins 


< > 

<1 
min 


NOT 

RECOMMENDED 

for 

OB 

scanning 


Recommended scanning time limits for these TIs 
(observe ALARA) 



Monitor TIS up to 10 weeks post-LMP, TIB thereafter. 

Figure 11.1. Graphical representation of the recommended exposure times at different index 
values for obstetric ultrasound, as listed in Table 11.1. 



11.3.1 EFSUMB(2009) 

Developments in real-time three-dimensional ultrasonic imaging have led to parents asking for 
souvenir (keepsake) video recordings of the foetus, sometimes at several stages during the pregnancy. 
An area of concern is the growth of services designed to provide such images and recordings without 
any diagnostic element to the scan. Often, such services are unable to provide counselling or offer 
guidance if signs of a foetal abnormality are unexpectedly revealed. Apart from such services, there 
many instances of diagnostic scans being prolonged in order to provide such recordings. 

Very little information is currently available regarding possible subtle biological effects of 
diagnostic levels of ultrasound on the developing human embryo or foetus, and the possibility 
of developmental effects in the brain cannot be ruled out. There is evidence that diagnostic levels 
of ultrasound can influence development of the brain in small animals, although it is not possible 
to extrapolate this finding to the human situation. A balance must always be maintained between 
diagnostic benefit and risk to the patient. Therefore, it is difficult to justify souvenir or keepsake 
scanning that has no diagnostic benefit. 

Recommendations: 

1. Ultrasound scans should not be performed solely for producing souvenir images or recordings 
of a foetus or embryo. 

2. The production of souvenir images or recordings for the parents to keep is reasonable if they 
are produced during a diagnostic scan, provided that this does not require the ultrasound 
exposure to be greater in time or magnitude (as indicated by the displayed MI and TI) than 
that necessary to produce the required diagnostic information. 

3. Attention is drawn to the recommendation of the EFSUMB Clinical Safety Statement for 
Diagnostic Ultrasound that ultrasound examinations should be performed only by competent 
personnel who are trained and updated in ultrasound safety matters. 




154 



Guidelines and recommendations 11 



11.3.2 WFUMB Policy Statement on Non-medical Use of Ultrasound 
(2008) 

The WFUMB disapproves of the use of ultrasound for the sole purpose of providing souvenir 
images of the foetus. Because the safety of an ultrasound examination cannot be assured, the use 
of ultrasound without medical benefit should be avoided. Furthermore, ultrasound should be 
employed only by health professionals who are well trained and updated in ultrasound clinical 
usage and bio-effects. 

11.3.2.1 WFUMB Recommendations on Non-medical Use of Ultrasound 

• The WFUMB disapproves of the use of ultrasound for the sole purpose of providing keepsake 
or souvenir images of the foetus. 

• Ultrasonography is a medical procedure that should only be carried out in the clinical setting 
where there is a medical indication and when carried out under the supervision of a physician 
or an expert. 

• The use of ultrasound to provide keepsake images or videos of the foetus may be acceptable if it 
is undertaken as part of the normal clinical diagnostic ultrasound examination, provided that 
it does not increase exposure to the foetus. 

• In the absence of supporting evidence of safety, caution should be used to minimize ultrasound 
exposure to the foetus. 

• When using ultrasound for non-medical reasons the ultrasound equipment display should be 
used to ensure that TI<0.5 and MI < 0.3. 

• Ultrasound imaging for non-medical reasons is not recommended unless carried out for 
education, training or demonstration purposes. 

• Live scanning of pregnant models for equipment exhibitions at ultrasound congresses is 
considered a non-medical practice that should be prohibited since it provides no medical benefit 
and affords potential risk to the foetus. 

11.3.3 ISUOG (2009) 

ISUOG disapproves of the use of ultrasound for the sole purpose of providing souvenir images 
of the foetus. There have been no reported incidents of human foetal harm in over 40 years 
of extensive use of medically indicated and supervised diagnostic ultrasound. Nevertheless, 
ultrasound involves exposure to a form of energy, so there is the potential to initiate biological 
effects. Some of these effects might, under certain circumstances, be detrimental to the 
developing foetus. Therefore, the uncontrolled use of ultrasound without medical benefit 
should be avoided. 

11.4 Conclusion 

Professional medical ultrasound society websites provide statements and guidelines 
addressing the continued safe practice of diagnostic ultrasound. These are regularly 
updated and should be consulted for the most up to date information. Many of these 
organizations also have rapid response groups whose job it is to provide informed 
comment when new safety issues arise. 

155 



11 Guidelines and recommendations 



Acknowledgement 

This chapter is a revised version of Chapter 11 in the second edition. The contribution of 
Stan Barnett to that chapter is acknowledged. 

References 

AIUM/NEMA. 1992. Standard for Real-time Display of Thermal and Mechanical Acoustic 
Output Indices on Diagnostic Ultrasound Equipment. Rockville, MD: American Institute 
of Ultrasound in Medicine. 

Barnett SB, ter Haar GR, Ziskin MC, Nyborg WL, Maeda K, Bang J. 1994. Current status of 
research on biophysical effects of ultrasound. Ultrasound Med Biol, 20, 205-218. 
Barnett SB, Rott HD, ter Haar GR, Ziskin MC, Maeda K. 1997. The sensitivity of biological 
tissue to ultrasound. Ultrasound Med Biol, 23, 805-812. 

Barnett SB, Abramowicz JS, Ziskin MC, Marsal K, Claudon C. 2010. Safety of nonmedical 
use of ultrasound. Ultrasound Med Biol, 36, 1209-1212. 

BMUS. 2010. Guidelines for the safe use of diagnostic ultrasound equipment. Ultrasound, 
18, 52-59. 

Brezinka C. 2010. Non-medical use of ultrasound in pregnancy; ethical issues, patient's 
rights and potential mis-use. Ultrasound Med Biol, 36, 1233-1236. 

Child SZ, Hartman CL, Schery LA, Carstensen EL. 1990. Lung damage from exposure to 
pulsed ultrasound. Ultrasound Med Biol, 16, 817-825. 

Duck FA, Henderson J. 1998. Acoustic output of modern ultrasound equipment: is it 
increasing? In Safety of Diagnostic Ultrasound, Barnett SB, Kossoff G (editors). New York, 
NY: Parthenon Publishing Group, pp. 15-26. 

EFSUMB. 1995. European Committee for Ultrasound Radiation Safety, Tutorial paper on 

transvaginal ultrasonography— safety aspects. Eur J Ultrasound, 1, 355-357. 

EFSUMB. 1996a. European Federation of Societies for Ultrasound in Medicine and 

Biology, Clinical safety statement for diagnostic ultrasound. Report from the European 

Committee for Ultrasound Radiation Safety. Eur } Ultrasound, 3, 283. 

EFSUMB. 1996b. Tutorial paper: thermal and mechanical indices. European Committee 

for Ultrasound Radiation Safety. Eur ] Ultrasound, 4, 145-150. 

FDA. 1985. 510(k) Guide for Measuring and Reporting Acoustic Output of Diagnostic 
Ultrasound. Rockville, MD: Food and Drug Administration, Centre for Devices and 
Radiological Health. 

FDA. 1993. Revised 510(k) Diagnostic Ultrasound Guidance for 1993. Rockville, MD: 
Centre for Devices and Radiological Health, US Food and Drug Administration. 
Henderson J, Whittingham TA, Dunn T. 1997. A review of the acoustic output of modern 
diagnostic ultrasound equipment. BMUS Bidl, 5, 10-14. 

IEC. 1992. Standard 1157: Requirement for the Declaration of Acoustic Output of Medical 
Diagnostic Equipment. Geneva, Switzerland: International Electrotechnical Commission. 
Martin K. 2010. The acoustic safety of new ultrasound technologies. Ultrasound, 18, 
110-118. 

Phillips RA, Stratmeyer M, Harris G. 2010. Safety and U.S. regulatory considerations in 
the non-medical use of medical ultrasound devices. Ultrasound Med Biol, 36, 1224-1228. 



156 



Guidelines and recommendations 11 



WFUMB. 1992. World Federation for Ultrasound in Medicine and Biology Symposium 
on Safety and Standardisation in Medical Ultrasound. Issues and recommendations 
regarding thermal mechanisms for biological effects of ultrasound, Barnett SB, Kossoff G 
(editors). Ultrasound Med Biol, 18, 731-814. 

WFUMB. 1998. World Federation for Ultrasound in Medicine and Biology Symposium on 
Safety of Ultrasound in Medicine: conclusions and recommendations on thermal and non- 
thermal mechanisms for biological effects of ultrasound, Barnett SB (editor). Ultrasound 
Med Biol, 24, 1-55. 

WFUMB. 2011. Statement on the Safe Use of Doppler Ultrasound During 11-14 week 
scans. Ultrasound Med Biol, 36, 1210. 

Whittingham TA. 2000. The acoustic output of diagnostic machines. In The Safe Use of 
Ultrasound in Medical Diagnosis, ter Haar G, Duck FA (editors). 2nd Edition. London, 
UK: The British Medical Ultrasound Society & The British Institute of Radiology, 
pp.16-31. 



157 



The Safe Use of Ultrasound in Medical Diagnosis 



Glossary 



Absorption coefficient: measure of the rate of decrease in the energy of an acoustic wave 
due to viscosity, thermal effects, chemical relaxation, etc., but excluding scattering. 
See also Attenuation coefficient 

Acoustic cavitation: formation and/or activity of gas-filled bodies in a medium exposed to 
an acoustic field. Commonly used with regard to cavitation associated with ultrasonic 
field as well 

Acoustic impedance: ratio of acoustic pressure to particle velocity in a medium exposed 
to an ultrasound wave, equal to the product of density and speed of sound 

Acoustic intensity: rate of acoustic energy flow through a unit area normal to the direction 
of wave propagation 

Acoustic streaming: the flow of fluid within an ultrasonic beam, in the direction of wave 
propagation: originally termed "quartz wind" 

AIUM: American Institute for Ultrasound in Medicine 

Attenuation coefficient: coefficient that describes the energy lost, by absorption 
and scattering, when an ultrasonic beam passes through a medium. The intensity 
absorption coefficient \i of a homogeneous medium is defined for a single frequency by 
the equation I = I 0 e - *™, where I 0 is the initial intensity and I is the intensity after the beam 
has travelled distance x. An amplitude attenuation coefficient a can also be defined, 
where f.i = 2a. Units: dB cm -1 , or Np crrr 1 

ASUM: Australasian Society for Ultrasound in Medicine 

BECA: beam calibrator; a device for quantifying acoustic fields. Also referred to as the 
NPL Ultrasound Beam Calibrator (UBC) 

BIR: The British Institute of Radiology 

BMUS: The British Medical Ultrasound Society 

Cavitation: see Acoustic cavitation 

Cavitation nucleus: point (e.g. impurity or structural irregularity in a liquid or soft tissue) 
from which a gas bubble may grow and oscillate under the action of an ultrasonic field 

CEN: Comite Europeen de Normalisation 

CENELEC: Comite Europeen de Normalisation Electrotechnique 

CFM: colour flow mapping using Doppler shift methods. Sometimes referred to as colour 
Doppler 

Collapse cavitation: see Inertia! cavitation 



159 



Glossary 

Contrast agents: agents that can be administered to patients to improve or enhance the 
diagnostic information in a scan specifically for ultrasound by differentially altering 
echogenicity; ultrasound contrast agents usually consist of a suspension of small gas 
bubbles stabilized against dissolution by the prescence of an encapsulating shell of 
lipid, protein or polymer 

Derating: process by which the attenuating effects of overlying structures in an ultrasonic 
beam are accounted for. See also in situ intensity 

ECURS: European Committee for Ultrasound Radiation Safety 

EFSUMB: European Federation of Societies for Ultrasound in Medicine and Biology 

Embryo: unborn offspring in first trimester of pregnancy 

ESWL: extracorporeal Shockwave lithotripsy 

Extravasation: leakage of blood cells through vessel walls 

FDA: Food and Drug Administration (USA) 

Foetus: unborn offspring in second or third trimester of pregnancy 

Fluid inertia: the tendency of a moving fluid to continue with a constant velocity 

Free radical: an atom or molecule having at least one unpaired electron and typically 
unstable and highly reactive. In animal tissue, free, radicals can damage cells and 
promote the progression of disease 

Gas bodies: accumulations of gas; examples include bubbles, intestinal gas and lung 
alveoli 

Hydrophone: device used for measuring acoustic pressure 
Hyperthermia: a temperatures above normal 
IEC: International Electrotechnical Commission 

in situ intensity: intensity at a target. This is usually calculated to take into account the 
attenuation of overlying tissues 

in vitro: literally "in glass". Used to refer to experiments carried out in the laboratory, in 
which samples can be studied in isolation from their host 

in vivo: used to refer to studies carried out in the intact living organism 

Inertial cavitation: activity of a gas body in an acoustic field characterized by rapid growth 
and subsequent collapse to a very small size, converting sound energy into heat, light, 
and shock waves. The motion during the collapse phase is governed by the inertia of 
the surrounding material rather than the acoustic and hydrodynamic pressure, hence 
"inertial". Previously referred to as collapse cavitation or transient cavitation 

Intensity: see Acoustic intensity 

I ob : output beam intensity; the spatial-average intensity at the transducer face 
I sala : spatial-average temporal-average intensity 
I : spatial-peak pulse-average intensity 
I ta : spatial-peak temporal-average intensity 
I. : temporal-average intensity 



160 



Lysis: cell disruption resulting from extracellular membrane damage 

Mechanical index: an output parameter related to the probability of an acoustic field 
giving rise to cavitation, used as an "on screen" label for the Output Display Standard 

MI: see Mechanical index 

Microstreaming: highly localized fluid movement in the vicinity of an oscillating gas 
body 

NCRP: National Council on Radiation Protection and Measurements (USA) 
NEMA: National Electrical Manufacturers Association (USA) 

Non-inertial cavitation: activity of a gas body in an acoustic field below the threshold 
acoustic pressure for inertial cavitation. Previously also known as stable cavitation. 

NPL: National Physical Laboratory (UK) 

ODS: see Output Display Standard 

Output Display Standard: AIUM/NEMA requirement for the display of safety-related 
indices on the screen of an ultrasonic scanner; the equivalent international standard 
promulgated by the IEC is "Medical electrical equipment. Part 2-37: Particular 
requirements for the safety of ultrasonic medical diagnosis and monitoring equipment", 
edition 2.0 

Peak negative pressure: see Peak rarefaction pressure 

Peak rarefaction pressure: maximum negative pressure amplitude in an ultrasonic wave 

Petechia: a very small spot of haemorrhage, typically on a surface such as skin or mucous 
membrane: also called petechial hemorrhage 

PPSI: pulse-pressure-squared integral; the integral of the square of the acoustic pressure 
over one pulse. 

prf : pulse repetition frequency 

PVDF: polyvinylidine fluoride; material used as the sensor in some hydrophones 

Radiation force: the force experienced by a solid object when it is placed in a progressive 
ultrasound wave. The force is directed along the beam in the direction of propagation 

Resonant bubble: a bubble in an ultrasonic field pulsating at one of its resonant 
frequencies; the term generally refers to the case for which the response consists of 
one radial maximum and one radial minimum per acoustic period. The amplitude of 
oscillation of a bubble is largest at its resonant frequency 

Resonant frequency: a frequency of vibration of an object determined by the physical 
characteristics of the object and characterized at low acoustic intensities by the response 
of the object being maximal 

Scattering coefficient: coefficient defining the portion of ultrasonic energy loss that is due 
to scattering. See Attenuation coefficient 

Sonoporation: ultrasonic induction of pores in the cellular membrane 

Spectral Doppler: a pulsed Doppler technique for blood flow waveform analysis 

Teratogenic effects: effects resulting in abnormal development of the embryo and foetus 



161 



Glossary 

Thermal index: an output parameter calculated as the ratio of attenuated acoustic power 
at a specified point to the attenuated acoustic power required to raise the temperature 
at that point in a specific tissue model by 1°C; used as an "on screen" label for the 
Output Display Standard 

TI: see Thermal index 

Transducer: source of ultrasound (also referred to as the "probe") 

Transducer self-heating: heating of an ultrasonic source (probe) due to dissipation of 
electrical energy within the probe itself 

Ultrasound bioeffects: the biological and physiological consequences of the passage of an 
ultrasonic wave through tissue 

WFUMB: World Federation for Ultrasound in Medicine and Biology 



162 



The Safe Use of Ultrasound in Medical Diagnosis 



Index 



A 

Absorption, 8, 19, 23, 28, 82 
Absorption coefficient, 9, 13, 15, 63 
Absorption coefficient of bone, 48 
Acoustic absorption coefficient, 48 
Acoustic cavitation, 91 
Acoustic frequencies, 82 
Acoustic impedance, 7, 16 
Acoustic intensity, 4, 82 
Acoustic output (or output) 

manufacturer declared values, 36-39 

regulation, 36, 42 

surveys, 33-40, 42 

trends, 39-40 

values, 33-40, 42 
Acoustic power (or power), 10 

measurement of, 19-20, 28, 33 
Acoustic pressure, 5 

measurement of, 19-21, 25-26, 33 

peak compression pressure, 19, 34 

peak rarefaction pressure, 18, 19, 21, 23, 
34-37, 39, 40, 42 
Acoustic radiation force impulse (ARFI), 

20,82 
Acoustic shock, 11 
Acoustic streaming, 20, 81, 83, 87 
AIUM/NEMA, 135, 138 
ALARA principle, 2 
Alveoli of the lung, 16 
Amniotic fluid, 83 
Attenuation, 8, 21, 32 
Attentuation coefficient, 7-9, 15, 82, 85 

B 

Backscattered energy, 9 
Bio-effects, 50, 81, 87, 88, 91 
Birth weight, 91 
Blanching of the choroid, 86 
Blood, 96 



Bone, 4, 10, 16, 94, 138 
Bone healing, 95 
Bone heating, 95 
Bone marrow, 51 

British Medical Ultrasound Society 
(BMUS), 151 

c 

Calibration 

hydrophone, 25, 26 

radiation force balance, 28 
Cardiac adverse effects, 109, 110, 118 
Cardiac response, 87 

Cavitation, 2, 14, 16, 18, 19, 22, 23, 37, 69-72, 

75, 81, 92, 93 
Cell density, 92 
Cell lysis, 91, 92 
Cell membrane, 88 
Cellular function, 94 
Cellular mechano-transduction, 88 
CE marking, 24 
Central nervous system, 63 
Choroid blanching, 87 
Chromosomal effects, 93 
Colour flow imaging, 47, 50 
Congenital heart defects, 54 
Contrast agent, 105-114, 116-118 
Contrast medium, 16 
Control settings 

effects on acoustic output, 31 

preset, 41 

worst case, 31, 32 

D 

Derated values, 21-22 
De-rating, 137 
Developmental effects, 86 
Diagnostic ultrasound, 105-118 
Divergent beam, 29 



163 



DNA, 93 
Doppler, 82, 148 
Doppler Ultrasound, 148 
Dose, 2 

Dose-response relationship, 59 
E 

EC, 135 

Echocardiography, 106, 109, 110, 

112-113, 117 
Efficiency, 15 
EFSUMB, 149, 154 
Elastography, 82 

Electrical earth leakage current, 135 

Electrical safety, 135 

Embryos, 97 

Epidemiology, 125-131 

Energy deposition, 10 

European Committee for Ultrasound 

Radiation Safety (ECURS), 145 
Exposure, 4 
Exposure duration, 54 
Exposure time, 152, 153 
Extracellular membrane, 93 

F 

FDA. See Food and Drug Administration 

Febrile obstetric patients, 51 

First trimester, 148 

Fluid-filled cysts, 84 

Fluid movement, 87 

Focused beam, 29 

Focusing, 10 

Foetal gestational age, 49 
Foetal neural tissue, 52 
Foetal weight reduction, 98 
Food and Drug Administration, 13, 18, 22, 
36-38, 42, 43, 134-136, 143 
510k, 136 
Free-field values, 21-22 
Free radical, 93 
Frequencies, 4, 5, 49 
Functional changes, 94 

G 

Gas, 8, 14 

Gene expression, 86 
Gestation, 52, 54 
in humans, 54 



Guidelines, 62, 151 

Guidelines and recommendations, 142 
H 

Haemodynamic shear, 88 
Haemorrhage, 86, 87 
Handedness, 127, 129-131 
Harmonics, 13 
Hazard, 31 

Healing of bone fractures, 86 
Heat conduction, 15 
Heating, 2, 4, 13, 46, 50 

tissue, 18, 19, 21-23 

transducer self heating, 22, 31 

transducer surface temperature, 23, 30, 32 
Heat shock proteins, 52 
Homeostatic processes, 51 
Hydrophone 

calibration, 25, 26 

measurement systems, 27-28 

membrane, 25, 26, 28, 33 

preamplifier, 25 

probe (needle), 26 
Hyperthermia, 51 

I 

IEC. See International Electrotechnical 

Commission 
IEC60601 Part 2-37, 136 
Imaging modes 

B-mode, 18, 30, 32, 34-42 

colour flow, 30, 32 

harmonic imaging, 33, 38, 42 

M-mode, 32 

pulsed Doppler, 21, 34-40, 42 
spectral Doppler, 18, 32, 33 

Index values, 152, 153 

In situ exposure, 11, 137 

Intensity, 10, 82 

measurements, 20-21, 26-27, 33 
spatial-average temporal-average 

intensity, 21 
spatial-peak pulse average intensity, 21 
spatial-peak temporal-average intensity, 18, 

21, 35, 40 

temporal-average intensity, 21, 26, 27, 32 
temporal-peak intensity, 19, 20, 32 
International Electrotechnical Commission, 

22, 23, 25, 26, 28, 30, 36, 134, 135 



164 



Index 



International guidelines, 146 
Intestine, 14, 16 
Intra-rectal probes, 136 
ISUOG, 155 



Longitudinal compressional wave, 5 
Longitudinal waves, 5, 16 
Lung, 14 
Lysosomes, 93 

M 

Maternal fever, 54 

Maternal hyperthermia, 54 

MDD. See Medical Devices Directive 

Mechanical effects, 13, 14, 69-72 

Mechanical index, 2, 11, 61, 71-73, 76, 134, 137, 

138, 145, 149 
definition, 22 
Mechanical process, 4 
Medical Devices Directive, 36, 

134, 140 
MHRA, 140 

MI. See Mechanical index 
Mitochondria, 93 
Mitosis, 92 

M-mode, 10, 49, 82, 138 
N 

National standards, 26, 29 
Neonatal ultrasound, 152 
Neper, 8 

Neuronal migration, 98 
Neurosensory responses, 86 
Non-linear acoustic effects, 4 
Non-linear enhancement, 82 
Non-linearity coefficient, 7 
Non-linear propagation, 25, 28, 42, 85 
Non-linear propagation effects, 11 
Non-medical use, 155 
Non-thermal effects, 147, 148 

o 

Obstetrics, 152 

ODS. See Output Display Standard 
Ophthalmology, 137 
Organogenesis, 88 
Output Display Standard, 48, 136, 
138, 145 



Peak rarefaction pressure, 4 
Permeability, 93 
Phantom 

thermal, 31, 41 
Physical effects, 86 
Piezoelectric transducer, 15 
Plane-wave assumption, 26 
Power, 82 
Preset 

scanner controls, 41 
Pressure ptilses, 12 
Propagation, 4 
Propagation speed, 7 
Protein synthesis, 94 
Pulsed Doppler, 10, 47, 49, 50, 63, 95 
Pulsed radiation force, 87 



Quality assurance (QA), 24 
R 

Radiation force, 70-71, 75, 81, 82 
Radiation force balance (RFB), 81 

calibration, 28 

target, 28-30 
Radiation force in fluids and tissues, 85 
Radiation force on soft tissue, 86 
Radiation pressure, 14 

Randomized controlled trials, 125, 126, 129-131 

Rarefaction pressure, 11 

Reflection, 7, 16 

Regulations, 22, 134 

Renal adverse effects, 109 

Reproductive integrity, 91, 92 

Risk, 143 

Risk assessment, 143 
Risk/benefit, 143 



Safety guidelines, 42 
Safety indices, 134 
Safety standards, 134 
Scanned transducer, 139 
Scanning time, 151 
Scatter, 8 
Scattering, 8, 9 
Scattering coefficient, 9 
Sensory effects, 86 



165 



Index 



Shear forces, 85 
Shear stress, 88 

Shear wave elasticity imaging (SWEI), 83 

Shear waves, 16 

Soft tissues, 4, 8, 96, 138 

Souvenir scanning, 151 

Spatial peak, pulse-average intensity, 137 

Spatial-peak temporal average intensity (J la ), 

10, 47, 137 
Specific heat, 15 
Spectral pulsed Doppler, 138 
Standing wave, 6 
Streaming, 83 
Surface waves, 16 

T 

Tactile sensation, 87 
Temperature increases, 46, 47, 49, 51, 63 
Temperature rise, 50, 138, 139 
Temperatures, 14 
tissue, 22, 23 

transducer face, 18, 23, 32 
Temperature thresholds, 57 
Temperature-time profiles, 59 
Teratogenic effects, 46, 54 
Thermal, 81 
Thermal bio-effects, 49 
Thermal conduction, 16 
Thermal dose, 56, 60, 61 
Thermal effects, 13, 46, 147 
Thermal exposures, 60 
Thermal index, 2, 60, 134, 137, 138, 145, 149 

bone thermal index (TIB), 23, 37, 38, 42 

cranial thermal index (TIC), 23, 37, 38, 42 

definition, 23 

soft tissue thermal index (TIS), 23, 37, 38, 42 
Thermally induced biological effects, 55 
Thermally induced teratogenic effects, 53 
Thermal phantom, 31, 41 
Thermal safety, 135 
Thresholds, 58 



TI. See Thermal index 
TIB, 138 
TIC, 138 

Time-averaged acoustic intensity, 15 
TIS, 138 

Tissue heating, 15 
Tissue perfusion, 15 
Total acoustic power, 4 
Track 1, 137 
Track 3, 137, 138, 143 
Transducer heating, 147 
Transducers, 15, 50 
Transducer self-heating 

measurement of, 31 

surface temperature, 31 

temperature limits, 23, 30 

temperature rise, 31 
Transducer surface temperatures, 134 
Transmission, 7 
Trans-vaginal probe, 136 
Transverse/shear wave, 5 

u 

Ultrasound contrast agents (UCA), 69, 70, 72, 

75, 76, 150 
Ultrasound dosimetry, 2 
Ultrasound heating, 48 
Ultrasound safety, 125, 126, 131 
Ultrasound wave propagation, 5 
Ultrastructural changes, 93 
Unscanned beam, 10, 82 
Unscanned transducer, 139 

V 

Vasculature, 96 
W 

Wavelength, 5, 6 
Wave propagation speed, 6 
WFUMB Safety Committee, 143-146 
Worst case values, 31, 34-36, 41 



166